Each electronic device that allows wireless communication incorporates a building block that is called an RF transceiver. It consists of a receiver and a transmitter. The basic function of a receiver is to down-convert high frequency signals to a lower intermediate frequency, to convert them into the digital domain, and to demodulate them. Some of the most challenging and key RF blocks can be found in the analog highfrequency part of a receiver, which is called the RF front-end. Therefore, the focus of this thesis and the core of research is related to the design and implementation of RF front-ends. Nevertheless, RF circuit design is a delicate process involving many time consuming steps. In industry whole design teams work on the radio. So, it is important to limit the scope. Particularly, multi-band LNAs, multi-band RF filters and mixers have been subjects of investigation. This thesis starts with an introductory part. It provides an overview of the existing standards for wireless communications and gives the state-of-the-art in multi-standard RF transceivers. The most important part of the introduction is related to the discussion about the trends in the field of wireless communications, IC technology and packaging as well as about their consequences. Based on these consequences the objectives of this thesis have been formulated. The primary goal is to investigate solutions, benefits, limitations and costs related to multi-standard operation of RF front-ends and their adaptivity to variable radio environments. The second goal is to highlight the optimization of RF front-ends that allows to achieve maximal performance with a certain power budget, while targeting full integration. The third goal is to investigate possibilities for low-voltage low-power circuit topologies in CMOS technology. Finally, the results and considerations presented in this thesis have to be validated through the design and implementation of the reconfigurable multi-band multi-standard DECT/Bluetooth RF front-end in 0.18 µm CMOS technology. At the end of the introductory part, the design flow of a multi-standard front-end has been introduced and the steps in this flow have been discussed. According to the multi-standard design flow, the selection of a suitable front-end architecture is the first step in the system level design. The double-quadrature low-IF front- end architecture with adaptive image rejection has been selected as the most suitable architecture for the DECT/Bluetooth multi-band front-end. The achievable maximal image rejection in the double-quadrature low-IF architectures is directly dependent on the performance of the RC polyphase filters. Therefore, in the next step the performance of broad-band RC polyphase filters has been investigated. The last step in the system level analysis has been to calculate RF and building block specifications. The calculation of RF specifications is straightforward, while the distribution of the building block specifications has been based on the analytical expressions for the voltage gain, noise figure and IIP3 of low-IF front-ends. It has not been possible to guarantee that this distribution will lead to a minimal power consumption or minimal occupied chip area of the on-chip front-end. Nevertheless, its purpose has been to provide the starting point for the circuit level design, where further design optimization has been done. In the double-quadrature low-IF front-end with adaptive image rejection, the performance of the down-conversion mixer has a great influence on the overall front-end performance. Actually, the down-conversion mixer can be considered as the most important building block in the double-quadrature down-converter. Therefore, the first step in the circuit level design and implementation has been to design and implement the folded switching mixer with current-reuse. The measurements results of this mixer have shown very good performance even at a supply voltage of 1 V. After the successful implementation of the folded switching mixer with current-reuse, the design and implementation of the reconfigurable multi-band DECT/Bluetooth complex mixer with adaptive image rejection has been the next step. The measured performance has been quite satisfactory, while the cost introduced due to the adaptive image rejection have been acceptable. The design and implementation of the reconfigurable multi-band DECT/Bluetooth LNA has been the last step towards the design and implementation of the reconfigurable multi-band DECT/Bluetooth RF front-end. An inductively-degenerated common-source LNA has been selected as the most-promising topology. The analysis of this LNA has been presented and a rather efficient design procedure yielding a minimum noise figure with a certain power consumption has been proposed. Finally, the evaluation of the cost-effectiveness of the designed reconfigurable multi-band DECT/Bluetooth RF front-end has been considered. It has been shown that the reconfigurable multi-band DECT/Bluetooth RF front-end with adaptive image rejection is an appropriate solution when an image rejection of more than 30 dB is required. At the end of the thesis some concluding remarks have been given. The most important are the following. First, the design and use of multi-standard RF transceivers is the way to increase hardware flexibility and functionality, as well as to improve the flexibility of set-makers on the market. Second, since the radio environment is variable, the application of adaptivity in RF transceivers results in the reduction of their power consumption. Third, the implementation of analog amplification at low supply voltages has to be based on inverters.
|Qualification||Doctor of Philosophy|
|Award date||9 Jan 2007|
|Place of Publication||Eindhoven|
|Publication status||Published - 2007|