Intelligent real-time environment and process adaptive radio frequency front-ends for ultra low power applications

Radio frequency;Adaptation;Low noise amplifier;LNA;Front-end;Fuzzy logic;Learning;Energy-per-bit;Mixer;Clustering;Process Variation;Channel

Banerjee, Debashis ; Chatterjee, Abhijit Electrical and Computer Engineering Wang, Hua Milor, Linda S. Kumar, Satish Cressler, John D. ; Chatterjee, Abhijit

University:Georgia Institute of Technology

Department:Electrical and Computer Engineering

关键词: Radio frequency;    Adaptation;    Low noise amplifier;    LNA;    Front-end;    Fuzzy logic;    Learning;    Energy-per-bit;    Mixer;    Clustering;    Process Variation;    Channel;   

Others  :  https://smartech.gatech.edu/bitstream/1853/53882/1/BANERJEE-DISSERTATION-2015.pdf

美国|英语

来源: SMARTech Repository

【 摘 要 】

In the thesis the design of process tolerant, use-aware radio-frequency front-ends were explored. First, the design of fuzzy logic and equation based controllers, which can adapt to multi-dimensional channel conditions, are proposed. Secondly, the thesis proves that adaptive systems can have multiple modes of operation depending upon the throughput requirements of the system. Two such modes were demonstrated: one optimizing the energy-per-bit (energy priority mode) and another achieving the lowest power consumption at the highest throughput (data priority mode). Finally, to achieve process tolerant channel adaptive operation a self-learning methodology is proposed which learns the optimal re-configuration setting for the system on-the-fly. Implications of the research are discussed and future avenues of further research are proposed.