Conflict-free joint decision by lag and zero-lag synchronization in laser network

TL;DR

The paper presents a photonic CMAB solution using a four-laser network that exploits lag synchronization to drive leader switching and zero-lag synchronization to prevent collisions, enabling conflict-free cooperative decisions. Through both Lang–Kobayashi-based simulations and a detailed experimental setup, the authors demonstrate high team rewards and low conflict rates in a 2-player, 2-slot scenario, with evidence of scalability to larger configurations. Comparisons against independent, non-cooperative configurations show the superiority of the cooperative approach, while discussions address practical stability and delay-matching challenges. The work suggests a viable path to scalable, fast, photonic decision-making systems leveraging chaotic laser dynamics.

Abstract

With the end of Moore's Law and the increasing demand for computing, photonic accelerators are garnering considerable attention. This is due to the physical characteristics of light, such as high bandwidth and multiplicity, and the various synchronization phenomena that emerge in the realm of laser physics. These factors come into play as computer performance approaches its limits. In this study, we explore the application of a laser network, acting as a photonic accelerator, to the competitive multi-armed bandit problem. In this context, conflict avoidance is key to maximizing environmental rewards. We experimentally demonstrate cooperative decision-making using zero-lag and lag synchronization within a network of four semiconductor lasers. Lag synchronization of chaos realizes effective decision-making and zero-delay synchronization is responsible for the realization of the collision avoidance function. We experimentally verified a low collision rate and high reward in a fundamental 2-player, 2-slot scenario, and showed the scalability of this system. This system architecture opens up new possibilities for intelligent functionalities in laser dynamics.

Figures (8)

• Figure 1: System architecture for addressing the competitive multi-armed (CMAB) problem in a 2-player, 2-slot context. a) Schematic diagram of decision-making using zero-lag synchronization in a laser network. b) Players' decision on slot machine selection facilitated by lag synchronization. The zero-lag synchronization enables conflict-free joint decision-making between Players 1 and 2.
• Figure 2: Numerical simulation results: about temporal waveforms and cross-correlation values. a) Temporal waveforms of laser intensities determined by the Lang-Kobayashi equations. b) Low-pass filtered temporal waveforms of laser intensity. c) Cross-correlation value between Lasers 1A and Laser 1B. d) Cross-correlation value between Lasers 1A and Laser 2B.
• Figure 3: Numerical simulation results: short-term cross-correlation values and decision-making. a) Short-term cross-correlation values $C_{1A}$ (blue curve), $C_{1B}$ (orange curve), $C_{2A}$ (yellow curve), and $C_{2B}$ (purple curve) calculated from Fig. \ref{['fig:simulation']}a. b) Enlarged view of short-term cross-correlation values. c) Accumulated rewards for Player 1 (blue curve), Player 2 (orange curve), and the team (black curve) in the decision-making. d) Conflict rate between Players 1 and 2 in decision-making.
• Figure 4: Experimental setup of the proposed decision-making system. ATT: voltage-controlled variable attenuator, CIRC: optical circulator, ISO: optical isolator, FC: fiber coupler, OSC: oscilloscope, PD: photodetector.
• Figure 5: Experimental results: temporal waveforms and cross-correlation values. a) Temporal waveforms of laser intensities. b) Low-pass filtered temporal waveforms of laser intensities. c) Cross-correlation value between Lasers 1A and Laser 1B ($\hat{C}_{1A,1B}(0)$). d) Cross-correlation value between Lasers 1A and Laser 2B ($\hat{C}_{1A,2B}(0)$).