Properties of Information Processing with Semiconductor Laser Networks

Semiconductor lasers (SLs) exhibit a wide range of dynamic behaviors under optical coupling and feedback, making them versatile systems for exploring complex dynamics [1]. Coupled SL topologies are particularly no-table for their ability to achieve synchronization [2], a property that underpins their potential for applications in brain-inspired information processing [3]. This study investigates the capabilities of SL networks (SLNs) for processing encoded information, using both numerical simulations and experimental implementations. By examining an all-to-all-coupled fiber network of SLs, we focus on how heterogeneity and synchronization influence the net-work's ability to transform and encode information. These insights mark a crucial step toward developing photonic integrated chips with SLN s, an alternative architecture approach for unconventional computing.