Keynote Speakers

Towards Physical Neural Networks for Wireless Communications

Marco Di Renzo – Prof. of Telecommunications Engineering at King’s College London

Abstract: A physical neural network is a type of artificial neural network in which an electrically adjustable material is used to emulate the function of a neural neuron model. The term “physical” neural network is used to emphasize the reliance on physical hardware utilized to emulate neurons as opposed to software-based approaches. In this talk, we discuss the role of physical neural networks in the context of wave-domain information processing for wireless communications. We will focus our attention on implementations based on reconfigurable metasurfaces, by considering the case of the recently proposed stacked intelligent metasurface technology.

Three Years in the Construction of the SKA-Low Telescope: Lessons Learned, Challenges and Future Perspectives

Lucio Tirone – Low AIV Lead Engineer @ SKAO

Abstract: After the first Level 1 Milestone was achieved last July 2025 (Array Assembly 0.5, or AA0.5), the full focus of the deployment of the SKA-Low Telescope is now on the next step, AA1. The scale of the Telescope is being quadrupled, passing from 4 to 16 fully populated Stations (each Station composed of 256 SKALA Antennas), and so does the complexity of the control system needed to operate it. New challenges are arising, from the consolidation of the novel technologies employed in the array, to the establishment of proper operation and maintenance procedures for the equipment already on the ground, to the planning for scaling up towards the future Milestone, AA2 in 2027, which will challenge the engineering and science teams with another staggering quadruplication, targeting 68 Stations. This upgrade will make SKA-Low the actual largest radio-telescope array on the planet, and will be the gateway to start the phase of Science Verification.

Optical fibers beyond communication

Roland Ryf – Department Head Advanced Photonics Research Nokia Bell Labs

Abstract: Digital signal processing (DSP) combined with coherent optical technology has revolutionized fiber-optic communication, delivering over an order of magnitude more capacity than conventional on-off keying systems. This approach significantly simplifies line system design by compensating for linear and partially nonlinear impairments in the digital domain rather than the physical domain. Additionally, it is a key enabler for novel high-capacity transmission systems based on multi-mode and multi-core fibers. 
Recently, similar techniques have been applied to reflective fiber sensing. This has resulted in distributed acoustic sensing (DAS) interrogators with enhanced sensitivity, longer reach, and compatibility with existing communication networks—effectively transforming current optical infrastructure into dense acoustic sensing arrays.