Sub-Terahertz Analog-to-Digital Conversion Using Integrated Soliton Microcombs

ended

Modern information and communication technology has been propelling the rapid expansion of signal spectrum bandwidth towards the level of hundreds of GHz and even 1 Terahertz. Such wideband analog signals produced in physical world must be converted to a stream of data bits via analog-to-digital conversion (ADC), for ultra-fast and flexible digital signal processing (DSP). However, the random electron fluctuations in semiconductors set a fundamental limitation on the performance of electronic (ADCs), leading to an inherent trade-off between the sampling accuracy and bandwidth. State-of-the-art electronic ADCs typically have only GHz-level analog bandwidth, which is becoming an increasingly severe limitation on high-speed DSP applications. Although the adoption of mode-locked lasers (MLLs) can overcome some limitations using the ultra-stable pulse train for precise time-domain sampling, the GHz-level repetition rate and the challenging integration of MLLs prevents any usability of photonics-assisted ADC in practical applications. In the CompADC project, I propose to develop a radically-new photonic ADC scheme using chip-scale dual optical frequency combs, enabling real-time digitization of ultra-wideband RF and microwave signals with a bandwidth of > 100 GHz. This envisaged performance is enabled by the emerging dissipative Kerr soliton (DKS) microcombs generated in SiN microresonators, which produces a new type of on-chip mode-locked emission of optical pulses with repetition rates exceeding 100 GH. These phase-locked dual microcombs (signal comb and local oscillator comb) will perform precise frequency-domain decomposition and parallel frequency down-conversion of ultra-wideband microwave signals to the detectable range of lower-speed electronics. This CompADC approach has the clear potential to offer unparalleled performance and chip-scale integration for modern ultra-wideband signal processing and communication applications.