Design Automation Algorithms Research On Field Coupled Nanocomputing

Ever since Gordon Moore proposed Moore’s Law in 1965,CMOS chips have become more integrated and more powerful.But the trend slowed after 2000.As CMOS technology approaches its physical limits,Moore’s Law will eventually fail.In this context,new computing devices based on nano technology usher in a wide range of development,field coupled nanocomputing(FCN)is one of the promising directions for its low power consumptions and heigh operation frequency.The uniqueness of FCN device lies in that the representation,calculation and transmission of information is all based on the field effect in local space,which means that the physical mechanism,circuit design,architecture is significantly different from the CMOS technology,resulting in a series of new theoretical and technical problems,among which FCN circuit design automation algorithms is an important research area.In order to achieve the FCN circuit design automation,the academia has put forward some specific algorithms for logic synthesis,placement and routing,and circuit simulation,but many of which are still the continuation of traditional CMOS design automation algorithm,and FCN circuit design automation algorithm research is still in its early development stage.The main work of this dissertation is a holistic study of FCN circuit design automation.Firstly,the main characteristics of logic synthesis,placement and routing and circuit simulation are summarized by analyzing previous works,and the ideas and routes of future research are clarified.Then,under the guidance of these observations,specific research is carried out respectively for the three sub-areas.The main research contents and innovations are summarized as follows:Firstly,the main characteristics of logic synthesis,placement and routing,and circuit simulation are summarized,and different research ideas and routes are sorted out according to the different characteristics of the three sub-fields.Secondly,in the logic synthesis part,the data structure of Majority Inverter Graph(MIG)is studied,and the realization scheme based on linear list and hash table is proposed.On this basis,a logic synthesis algorithm with depth optimization function is proposed.Secondly,in the logic synthesis part,we focus on the Majority Inverter Graph(MIG)data structure,and its application in logic representation,logic transformation and logic optimization.In logic representation,a MIG implementation scheme based on linear list and hash table and a new text format are proposed.In logic transformation,common transformation operators in network creation stage and logic optimization stage are implemented respectively.In logic optimization,a logic synthesis algorithm orientated on depth optimization is proposed.Thirdly,in the part of placement and routing,starting from the formal definition of the problem,a solution space searching problem of placement and routing is proposed.Then two search algorithm frameworks are proposed for the different solution space types of hierarchical and random placements.The first one is a tree solution space deterministic search algorithm framework for hierarchical placement,and the effectiveness of this framework is verified by taking the generalized depth-first search algorithm as an example.The second is an intelligent search algorithm framework for the complete solution space for random placement,and the combination of genetic algorithm and A*algorithm is taken as an example to verify its effectiveness.Finally,the concept of Design Rule Check(DRC)for FCN circuits is proposed and embedded into the placement and routing algorithm,which achieves good cellular layout effect.Fourthly,in the circuit simulation part,taking physical model and calculation method as the core,several important simulation models of FCN circuit are introduced,and a circuit simulation framework is devised.It is applied to energy analysis of quantum-dot cellular automata(QCA)and circuit simulation of nanomagnetic logic(NML),where QCA and NML are two different physical implementation schemes of FCN.Based on the previous results,a graphical interface tool for designing and simulating FCN circuits is developed.