In order to address the challenge of increasing data rates, next generation optical communication networks will require the co-integration of electronics and photonics. Heterogeneous integration of these technologies has shown promise, but will eventually become bandwidth limited. Faster monolithic approaches will, therefore, be needed, but monolithic approaches using complementary metal-oxide-semiconductor (CMOS) electronics and silicon photonics are typically limited by their underlying electronic or photonic technologies. Here, we report a monolithically integrated electro-optical transmitter that can achieve symbol rates beyond 100 GBd. Our approach combines advanced bipolar CMOS with silicon plasmonics, and addresses key challenges in monolithic integration through the co-design of the electronic and plasmonic layers, including thermal design, packaging, and a nonlinear organic electro-optic material. To illustrate the potential of our technology, we develop two modulator conceptsan ultra-compact plasmonic modulator and, alternatively, a silicon-plasmonic modulator with photonic routing -both directly processed onto the bipolar CMOS electronics.

The seamless integration of III-V nanostructures on silicon is a long-standing goal and an important step towards integrated optical links. In the present work, we demonstrate scaled and waveguide coupled III-V photodiodes monolithically integrated on Si, implemented as InP/In0.5Ga0.5As/InP p-i-n heterostructures. The waveguide coupled devices show a dark current down to 0.048 A/cm2 at −1 V and a responsivity up to 0.2 A/W at −2 V. Using grating couplers centered around 1320 nm, we demonstrate high-speed detection with a cutoff frequency f3dB exceeding 70 GHz and data reception at 50 GBd with OOK and 4PAM. When operated in forward bias as a light emitting diode, the devices emit light centered at 1550 nm. Furthermore, we also investigate the self-heating of the devices using scanning thermal microscopy and find a temperature increase of only ~15 K during the device operation as emitter, in accordance with thermal simulation results.

A deep learning (DL) based digital backpropagation (DBP) method with a 1 dB SNR gain over a conventional 1 step per span DBP is demonstrated in a 32 GBd 16QAM transmission across 1200 km. The new DL-DPB is shown to require 6 times less computational power over the conventional DBP scheme. The achievement is possible due to a novel training method in which the DL-DBP is blind to timing error, state of polarization rotation, frequency offset and phase offset. An analysis of the underlying mechanism is given. The applied method first undoes the dispersion, compensates for nonlinear effects in a distributed fashion and reduces the out of band nonlinear modulation due to compensation of the nonlinearities by having a low pass characteristic. We also show that it is sufficient to update the elements of the DL network using a signal with high nonlinearity when dispersion or nonlinearity conditions changes. Lastly, simulation results indicate that the proposed scheme is suitable to deal with impairments from transmission over longer distances.

Resonant modulators encode electrical data onto wavelength-multiplexed optical carriers. Today, silicon microring modulators are perceived as promising to implement such links; however, they provide limited bandwidth and need thermal stabilization systems. Here we present plasmonic micro-racetrack modulators as a potential successor of silicon microrings: they are equally compact and compatible with complementary-metal–oxide–semiconductor-level driving voltages, but offer electro-optical bandwidths of 176 GHz, a 28 times improved stability against operating temperature changes and no self-heating effects. The temperature-resistant organic electro-optic material enables operation at 85 °C device temperature. We show intensity-modulated transmission of up to 408 Gbps at 12.3 femtojoules per bit with a single resonant modulator. Plasmonic micro-racetrack modulators offer a solution to encode high data rates (for example, the 1.6 Tbps envisioned by next-generation communications links) at a small footprint, with low power consumption and marginal, if no, temperature control.

A multiple-access noise-based frequency offset modulation (N-FOM) system is studied in the context of near-far problem. A closed-form expression for the signal-to-noise ratio is derived and the bit error rate (BER) is evaluated. The result shows that the near-far problem compromises the performance of the multiple-access N-FOM system. Performance comparison with a standard spread-spectrum system indicates that the nearfar problem has an inherently different effect on the N-FOM communication. The error rate of a node in the N-FOM system increases at a much higher rate as the number of simultaneous communication links increases. However for small number of simultaneous links, an adaptive power control scheme can help minimize the threat of the near-far problem.