Research on the stochastic behavior of traffic flow is important to understand the intrinsic evolution rules of a traffic system. By introducing an interactional potential of vehicles into the randomization step, an improved cellular automata traffic flow model with variable probability of randomization is proposed in this paper. In the proposed model, the driver is affected by the interactional potential of vehicles before him, and his decision-making process is related to the interactional potential. Compared with the traditional cellular automata model, the modeling is more suitable for the driver's random decision-making process based on the vehicle and traffic situations in front of him in actual traffic. From the improved model, the fundamental diagram(flow–density relationship) is obtained, and the detailed high-density traffic phenomenon is reproduced through numerical simulation.
As feature sizes shrink,low energy consumption,high reliability and high performance become key objectives of network-on-chip(NoC) design.In this paper,an integrated approach is presented to map IP cores onto NoC architecture and assign voltage levels for each link,such that the communication energy is minimized under constraints of bandwidth and reliability.The design space is explored using tabu search.In order to select optimal voltage level for the links,an energy-efficiency driven heuristic algorithm is proposed to perform energy/reliability trade-off by exploiting communication slack.Experimental results show that the ordinary energy optimization techniques ignoring the influence of voltage on fault rates could lead to drastically decreased communication reliability of NoCs,and the proposed approach can produce reliable and energy-efficient implementations.
