T. Tanabe, T. Oya
Graduate school of engineering, Yokohama National University, Japan
pp. 179 - 183
Keywords: single-electron, information proccessing, nature-inspired, nonliner-problem, nanoelectronics
We propose a new single-electron (SE) circuit that shows a unique information-processing by mimicking a foraging behavior of a swarm of honeybees. Recently, a “nature-inspired” or “biomimetic” technique based on natural world phenomena or biological behaviors has been proposed for the purpose of developing novel functional systems using emerging nano-scaled devices that process information. In this study, we focus on a foraging behavior of a swarm of honeybees as the nature-inspired information-processing to apply a SE circuit that is the targeted device in this study. It has been reported that the foraging behavior of the honeybees can be assumed to be a certain type of information processing that is the solving nonlinear problems. For example, they fly around in random to discover nectar sources firstly. They return to the nest after discovery. Secondly, they share information about the place, quality, and the amount of the nectar sources in the nest by a “waggle dance.” Finally, after the sharing the information, appropriate numbers of honeybees goes for the better nectar sources. By repeating the process, they optimize the process. To materialize an “artificial honeybee circuit,” there are at least three important points, i.e., the “random flying (walk)” to find foods, the “sharing information” in the swarm, and the “updating information by repeating process,” should be focused. To mimic the random walk behavior firstly, we utilize a two-dimensional arrayed SE oscillator (SEO) system. Because the system can generate spatiotemporal patterns, i.e., “voltage waves,” and the pattern propagates with randomness on the system. Here, we assume the propagating random voltage waves on the system to be the random walk of our electronic honeybees. The general single-electron circuit that is a certain type of a quantum device can control an individual electron by controlling a quantum effect. A main component of the circuit is a tunneling junction. And the electron can pass through the junction under a special condition, i.e., the junction has a threshold value (voltage across it) for the electron to tunnel. The SEO that is main component of the system consists of a bias voltage source, a high resistance, and the tunneling junction in series. It operates as a threshold element because of the contained tunneling junction. As the next step, we try to mimic the “sharing information” and the “updating information.” We consider that the memory circuits are required because the honeybees memorize kinds of information to forage. In this study, we utilize single-electron memory (SEM) circuits. It consists of a bias voltage source, a capacitor, and two tunneling junctions in series. For our honeybees, the SEMs that are prepared as same numbers as the SEOs are arrayed as same as the two-dimensional SEO system. By connecting the layers of the SEOs and the SEMs each other appropriately, our system can operate as desired. By using a Monte Carlo simulation for the simple test circuit, we confirmed that our circuit operated effectively like the real honeybees. Therefore, the “honeybee inspired single-electron circuit” can show unique information-processing.