Periodic Reporting for period 2 - TeraSlice (Terahertz Analogue-to-Digital Conversion Using Photonic Chipscale Soliton Frequency Combs and Massively Parallel Spectrally Sliced Detection)
Reporting period: 2021-01-01 to 2023-06-30
High-speed digital signal processing (DSP) has seen tremendous performance increases over the last years, primarily driven by massive parallelization of logic operations in large-scale CMOS circuits that operate at comparably low clock frequencies. This has led to digital processors that would in principle allow for real-time digital processing of ultra-broadband signals with analogue bandwidths of hundreds of GHz already today. Acquisition of such signals, however, is still impossible due to limited bandwidth scalability of conventional analogue-to-digital converters (ADC). Within TeraSlice, we will explore and demonstrate concepts that can overcome these limitations by exploiting photonically assisted spectral parallelization of ADC interfaces, thereby enabling conversion of waveforms with bandwidths in excess of 200 GHz with the potential for further scalability beyond 1 THz.
The TeraSlice approach has the potential to disrupt a wide field of highly relevant applications, ranging from high-speed communications to metrology and sensing and further to high-field electron paramagnetic resonance (EPR) spectroscopy.
TeraSlice has the following overall scientific and technological objectives:
(i) to establish unprecedentedly compact and efficient Kerr soliton frequency comb generators for realizing an ultra-fast photonic ADC
(ii) to establish a novel class of high speed electro-optic modulators using capacitively-coupled silicon-organic hybrid (SOH) devices
(iii) to establish a new paradigm for Terahertz photonic analogue-to-digital conversion based on integrated photonics and soliton Kerr frequency combs
The TeraSlice approach has the potential to disrupt a wide field of highly relevant applications, ranging from high-speed communications to metrology and sensing and further to high-field electron paramagnetic resonance (EPR) spectroscopy.
TeraSlice has the following overall scientific and technological objectives:
(i) to establish unprecedentedly compact and efficient Kerr soliton frequency comb generators for realizing an ultra-fast photonic ADC
(ii) to establish a novel class of high speed electro-optic modulators using capacitively-coupled silicon-organic hybrid (SOH) devices
(iii) to establish a new paradigm for Terahertz photonic analogue-to-digital conversion based on integrated photonics and soliton Kerr frequency combs
EPFL kicked off the project on February 14, 2020 with all consortium partners in Lausanne, Switzerland. It was responsible for the entire scientific and administrative managements of the project.
THALES has worked on the conceptual system design of the demonstrator, and defined the complete architecture and theoretical model. It has also performed additional simulations predicting key performance; such as signal to noise ratio (SNR), effective number of bits (ENOB) and dynamic range (DR).
Chip-level building blocks are key to demonstrating the TeraSlice signal processing concept. The first-generation CC-SOH modulators were internally fabricated and characterized at KIT. The results have been accepted for publication. Detector arrays and spectral slicing filters were designed with support by LGT and external foundries.
EPFL focused on developing a local oscillator consisting of a compact soliton microcomb with very low noise. EPFL fabricated the first generation of silicon nitride microresonators for TeraSlice with support by LGT, providing the required specifications for the TeraSlice comb source, and successfully generated a single soliton microcomb. It further studied the phase noise performance of the self-injection locked lasers, the soliton microcomb repetition rate and the nonlinear dynamics that govern the phase noise transduction between the pump and the comb lines.
VANGUARD focused on the optical interconnects for the multi-chip assembly within TeraSlice. The first generation of optical interfaces and the corresponding photonic wire bonding process was developed. The individual interfaces can now be arbitrarily connected with each other for the TeraSlice assembly, using the photonic wire bonding process.
CNIT and KIT have progressed towards the development and implementation of the dedicated DSP software for slice stitching. This is part of the full test performance of the packaged ADC system.
All members engaged with different audiences to raise awareness of TeraSlice. CNIT designed and implemented the Teraslice website which was officially launched in April 2020 [https://www.teraslice-fetopen.eu/].
THALES has worked on the conceptual system design of the demonstrator, and defined the complete architecture and theoretical model. It has also performed additional simulations predicting key performance; such as signal to noise ratio (SNR), effective number of bits (ENOB) and dynamic range (DR).
Chip-level building blocks are key to demonstrating the TeraSlice signal processing concept. The first-generation CC-SOH modulators were internally fabricated and characterized at KIT. The results have been accepted for publication. Detector arrays and spectral slicing filters were designed with support by LGT and external foundries.
EPFL focused on developing a local oscillator consisting of a compact soliton microcomb with very low noise. EPFL fabricated the first generation of silicon nitride microresonators for TeraSlice with support by LGT, providing the required specifications for the TeraSlice comb source, and successfully generated a single soliton microcomb. It further studied the phase noise performance of the self-injection locked lasers, the soliton microcomb repetition rate and the nonlinear dynamics that govern the phase noise transduction between the pump and the comb lines.
VANGUARD focused on the optical interconnects for the multi-chip assembly within TeraSlice. The first generation of optical interfaces and the corresponding photonic wire bonding process was developed. The individual interfaces can now be arbitrarily connected with each other for the TeraSlice assembly, using the photonic wire bonding process.
CNIT and KIT have progressed towards the development and implementation of the dedicated DSP software for slice stitching. This is part of the full test performance of the packaged ADC system.
All members engaged with different audiences to raise awareness of TeraSlice. CNIT designed and implemented the Teraslice website which was officially launched in April 2020 [https://www.teraslice-fetopen.eu/].
Photonically assisted ADC are – despite decades of research and publications – still in their infancy and currently do not serve any technical applications. Nevertheless, the strong need for high-speed signal acquisition in various applications such as high-speed wireless communications beyond 5G or secure wireless transmission indicates the growing importance and role that photonic ADC can play in the future. The TeraSlice approach will overcome the shortcomings of previous photonic ADC concepts and hence, has the clear potential not only to enter, but also to disrupt technical applications. Specifically, the project will provide the technological foundation for a paradigm shift, which is enabled by a photonic ADC concept with unprecedented bandwidth scalability and fully amenability to chip-scale integration, and which will lay the foundations for a future emerging technology.
Acquisition of large-bandwidth signals in the 0 … 320 GHz domain is of dramatic interest for a variety of applications. These applications comprise areas of high commercial impact, e.g. ultra-broadband wireless communications within and beyond 5G and Internet-of-Things networks or security applications related to next-generation radar systems and secure wireless transmission. In this context, THALES explores the application potential of the TeraSlice approach for the analysis of radio frequency (RF) spectrum occupation, also referred to as spectrum sensing, which is of highest practical interest in several fields. Similar challenges exist in the area of high-performance radar systems that are addressed both by CNIT and THALES.
TeraSlice will enable photonic ADC that are disruptive, both with respect to performance and to viability of fully integrated chip-scale systems, and that provide a cost-effective approach to ultra-broadband signal processing. The impact of TeraSlice will unlock new markets with substantial economic potential worldwide. In fact, given the unique expertise and the strong international position of the participating research groups and companies, TeraSlice has the clear potential to reinforce and establish European leadership in the field of high-speed photonic-electronic signal processing.
Acquisition of large-bandwidth signals in the 0 … 320 GHz domain is of dramatic interest for a variety of applications. These applications comprise areas of high commercial impact, e.g. ultra-broadband wireless communications within and beyond 5G and Internet-of-Things networks or security applications related to next-generation radar systems and secure wireless transmission. In this context, THALES explores the application potential of the TeraSlice approach for the analysis of radio frequency (RF) spectrum occupation, also referred to as spectrum sensing, which is of highest practical interest in several fields. Similar challenges exist in the area of high-performance radar systems that are addressed both by CNIT and THALES.
TeraSlice will enable photonic ADC that are disruptive, both with respect to performance and to viability of fully integrated chip-scale systems, and that provide a cost-effective approach to ultra-broadband signal processing. The impact of TeraSlice will unlock new markets with substantial economic potential worldwide. In fact, given the unique expertise and the strong international position of the participating research groups and companies, TeraSlice has the clear potential to reinforce and establish European leadership in the field of high-speed photonic-electronic signal processing.