CN106027032A - RM logic circuit delay optimization method in unit delay model - Google Patents

RM logic circuit delay optimization method in unit delay model Download PDF

Info

Publication number
CN106027032A
CN106027032A CN201610341295.8A CN201610341295A CN106027032A CN 106027032 A CN106027032 A CN 106027032A CN 201610341295 A CN201610341295 A CN 201610341295A CN 106027032 A CN106027032 A CN 106027032A
Authority
CN
China
Prior art keywords
delay
tree
forest
simplest
logic
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201610341295.8A
Other languages
Chinese (zh)
Inventor
王翔
何振学
肖利民
谷飞
李明哲
苏书宾
霍志胜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beihang University
Original Assignee
Beihang University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beihang University filed Critical Beihang University
Priority to CN201610341295.8A priority Critical patent/CN106027032A/en
Publication of CN106027032A publication Critical patent/CN106027032A/en
Pending legal-status Critical Current

Links

Classifications

    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03KPULSE TECHNIQUE
    • H03K19/00Logic circuits, i.e. having at least two inputs acting on one output; Inverting circuits
    • H03K19/20Logic circuits, i.e. having at least two inputs acting on one output; Inverting circuits characterised by logic function, e.g. AND, OR, NOR, NOT circuits

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Design And Manufacture Of Integrated Circuits (AREA)

Abstract

一种单位延时模型下RM逻辑电路延时优化方法,其特征在于:该方法具体步骤包括:步骤1,读入布尔Boolean逻辑电路;步骤2,利用RM表达式化简方法得到含与项数最少的最简RM逻辑表达式;步骤3,基于哈夫曼Huffman树构造算法对最简RM逻辑表达式中的每个与项进行延时分解,使得每个与项的延时最小;步骤4,基于Huffman树构造算法对由所有与项组成的最简RM逻辑表达式进行延时分解,使得最简RM逻辑表达式的延时最小;步骤5,输出最简RM逻辑表达式的最小延时。本发明提供的延时优化方法,可快速有效地将布尔逻辑电路转换为具有最小延时的RM逻辑电路,进而降低电路延时、提高电路的工作速度和性能。

A method for optimizing the time delay of an RM logic circuit under a unit time delay model, characterized in that: the specific steps of the method include: Step 1, reading in a Boolean logic circuit; Step 2, using the RM expression simplification method to obtain the number of items containing The least simplest RM logic expression; Step 3, based on the Huffman Huffman tree construction algorithm, decompose the time delay of each AND item in the simplest RM logic expression, so that the delay of each AND item is the smallest; Step 4 , based on the Huffman tree construction algorithm, decompose the delay of the simplest RM logic expression composed of all AND items, so that the delay of the simplest RM logic expression is the smallest; step 5, output the minimum delay of the simplest RM logic expression . The delay optimization method provided by the invention can quickly and effectively convert the Boolean logic circuit into the RM logic circuit with the minimum delay, thereby reducing the circuit delay and improving the working speed and performance of the circuit.

Description

一种单位延时模型下RM逻辑电路延时优化方法A Delay Optimization Method for RM Logic Circuit under Unit Delay Model

【技术领域】【Technical field】

本发明涉及里德穆勒Reed-Muller(RM)逻辑电路优化方法,尤其涉及一种单位延时模型下Reed-Muller逻辑电路延时优化方法。属于逻辑电路综合优化技术领域。The invention relates to a Reed-Muller (RM) logic circuit optimization method, in particular to a Reed-Muller logic circuit delay optimization method under a unit delay model. The invention belongs to the technical field of logic circuit synthesis optimization.

【背景技术】【Background technique】

数字电路既可以由基于AND/OR/NOT(与/或/非)运算的Boolean(布尔)逻辑实现,也可以由基于AND/XOR(与/异或)或OR/XNOR(或/同或)运算的Reed-Muller(RM)逻辑实现。对于异或运算较为频繁的算术电路、奇偶校验电路和通信电路等电路而言,与Boolean逻辑实现形式相比,RM逻辑实现形式在功耗、面积和速度等方面具有较大的优势。此外,异或门某一输入的变化会直接引起其输出的变化,所以RM逻辑电路也具有较好的可测试性。RM逻辑的这些特性已引起人们的广泛关注,并已成为逻辑电路设计领域的研究热点。Digital circuits can be implemented by Boolean (Boolean) logic based on AND/OR/NOT (and/or/not) operations, or by AND/XOR (and/exclusive or) or OR/XNOR (or/same or) Reed-Muller (RM) logic implementation of the operation. For circuits such as arithmetic circuits, parity check circuits, and communication circuits with frequent XOR operations, compared with Boolean logic implementations, RM logic implementations have greater advantages in terms of power consumption, area, and speed. In addition, the change of a certain input of the XOR gate will directly cause the change of its output, so the RM logic circuit also has better testability. These properties of RM logic have attracted widespread attention and become a research hotspot in the field of logic circuit design.

随着集成电路的快速发展,单个芯片上集成的晶体管的数量以及金属互连线的层数都在不断增长,这些因素导致电路延时在不断增长。因此,延时已经成为集成电路设计的一个重要优化目标。特别对于对工作速度要求较高的高速超大规模集成电路来说,延时优化也已成为高速超大规模集成电路设计的重要组成部分。此外,由于组合电路延时决定时序电路循环周期的下界,因此,组合电路的延时优化已受到电路设计人员的普遍关注。With the rapid development of integrated circuits, the number of transistors integrated on a single chip and the number of layers of metal interconnection lines are constantly increasing, and these factors lead to continuous increase of circuit delay. Therefore, delay has become an important optimization goal in IC design. Especially for high-speed VLSI that requires a higher working speed, delay optimization has also become an important part of high-speed VLSI design. In addition, since the delay of combinational circuits determines the lower bound of the cycle period of sequential circuits, the delay optimization of combinational circuits has been widely concerned by circuit designers.

然而,现有针对RM逻辑电路的优化方法主要集中在功耗优化、面积优化以及功耗与面积协同优化,而对RM逻辑电路延时优化的研究相对匮乏。此外,现有针对RM逻辑电路延时优化的研究具有较大的局限性且优化效率较低。因此,开展RM逻辑电路延时优化方法研究是对RM逻辑电路优化方法体系的重 要补充,对集成电路优化设计具有重要意义。在电路逻辑设计时,通常忽略由导线和电容等因素造成的延时,而只考虑基本门延时。由于单位延时模型也广泛用于集成电路的延时估计,因此本发明采用单位延时模型来优化RM逻辑电路延时。本发明提供的一种单位延时模型下RM逻辑电路延时优化方法可获得具有最小延时的RM逻辑电路,因此本发明可用于逻辑电路的综合延时优化。本发明提供的一种单位延时模型下RM逻辑电路延时优化方法是对RM逻辑电路优化方法体系的重要补充,是对当前以Boolean逻辑为主的集成电路逻辑级自动优化设计方法学的重要完善。However, existing optimization methods for RM logic circuits mainly focus on power optimization, area optimization, and co-optimization of power consumption and area, while research on delay optimization of RM logic circuits is relatively scarce. In addition, the existing research on delay optimization of RM logic circuits has relatively large limitations and low optimization efficiency. Therefore, research on the delay optimization method of RM logic circuit is an important supplement to the optimization method system of RM logic circuit, and it is of great significance to the optimal design of integrated circuits. In circuit logic design, the delay caused by factors such as wires and capacitance is usually ignored, and only the basic gate delay is considered. Since the unit delay model is also widely used in delay estimation of integrated circuits, the present invention uses the unit delay model to optimize the delay of RM logic circuits. The RM logic circuit delay optimization method under the unit delay model provided by the invention can obtain the RM logic circuit with the minimum delay, so the invention can be used for comprehensive delay optimization of logic circuits. The RM logic circuit delay optimization method under a unit delay model provided by the present invention is an important supplement to the RM logic circuit optimization method system, and is an important contribution to the current logic-level automatic optimization design methodology of integrated circuits based on Boolean logic. Complete.

【发明内容】【Content of invention】

为解决上述问题,本发明提供了一种单位延时模型下RM逻辑电路延时优化方法。本方法首先对RM逻辑表达式进行化简,得到含与项数最少的最简RM逻辑表达式;然后基于Huffman(哈夫曼)树构造算法对最简RM逻辑表达式中每个与项进行延时分解,使得每个与项的延时最小;最后基于Huffman树构造算法对最简RM逻辑表达式进行延时分解,使得最简RM逻辑表达式的延时最小。In order to solve the above problems, the present invention provides a delay optimization method for RM logic circuits under the unit delay model. This method first simplifies the RM logic expression to obtain the simplest RM logic expression with the least number of AND items; then based on the Huffman (Huffman) tree construction algorithm, each AND item in the simplest RM logic expression is obtained. Delayed decomposition minimizes the delay of each AND term; finally, the simplest RM logic expression is decomposed based on the Huffman tree construction algorithm to minimize the delay of the simplest RM logic expression.

具体来说,本发明提供了一种单位延时模型下RM逻辑电路延时优化方法,该方法具体步骤包括:Specifically, the present invention provides an RM logic circuit delay optimization method under a unit delay model, and the specific steps of the method include:

步骤1,读入Boolean逻辑电路;Step 1, read in the Boolean logic circuit;

步骤2,利用RM表达式化简方法(如代数化简法、无关项化简法等)得到含与项数最少的最简RM逻辑表达式;Step 2, using the RM expression simplification method (such as algebraic simplification method, irrelevant item simplification method, etc.) to obtain the simplest RM logic expression containing the least number of terms;

步骤3,基于Huffman树构造算法对最简RM逻辑表达式中的每个与项进行延时分解,使得每个与项的延时最小;Step 3, based on the Huffman tree construction algorithm, decompose the delay of each AND item in the simplest RM logic expression, so that the delay of each AND item is the smallest;

步骤4,基于Huffman树构造算法对由所有与项组成的最简RM逻辑表达式进行延时分解,使得最简RM逻辑表达式的延时最小;Step 4, based on the Huffman tree construction algorithm, perform delay decomposition on the simplest RM logic expression composed of all AND items, so that the delay of the simplest RM logic expression is the smallest;

步骤5,输出最简RM逻辑表达式的最小延时。Step 5, output the minimum delay of the simplest RM logic expression.

其中,步骤3包括:Among them, step 3 includes:

步骤31,计算与项中含有的输入变量数n;Step 31, calculating the number n of input variables contained in the AND item;

步骤32,若输入变量数n等于1,则将该与项的延时设置为该输入变量的延时;若输入变量数n大于1,则首先为n个输入变量创建n个叶子结点,其次根据d1,d2,...,di,...,dn(1≤i≤n,di代表第i个输入变量的延时,dn代表第n个输入变量的延时)构建具有n棵二叉树的森林F={T1,T2,...,Ti,...Tn},1≤i≤n,其中每棵二叉树Ti仅具有一个延时为di的结点,最后重复执行以下操作直到森林F中仅具有一棵二叉树(即Huffman树):从森林F中选择根结点延时最小的两棵树l和r;创建一棵以l作为左子树并以r作为右子树的新树,其中,新树根结点的延时等于两子树延时中较大的加1;从森林F中删除被选择的两棵树l和r;将创建的新树加入森林F。最后将该与项的延时设置为Huffman树根结点的延时。Step 32, if the number of input variables n is equal to 1, set the delay of the AND item as the delay of the input variable; if the number of input variables n is greater than 1, first create n leaf nodes for n input variables, Secondly, according to d 1 ,d 2 ,...,d i ,...,d n (1≤i≤n, d i represents the delay of the i-th input variable, and d n represents the delay of the n-th input variable When) to build a forest F={T 1 , T 2 ,...,T i ,...T n } with n binary trees, 1≤i≤n, where each binary tree T i has only one delay d i node, and finally repeat the following operations until there is only one binary tree (ie Huffman tree) in the forest F: select two trees l and r with the smallest root node delay from the forest F; create a tree with l A new tree with r as the left subtree and r as the right subtree, where the delay of the root node of the new tree is equal to the greater of the delays of the two subtrees plus 1; delete the selected two trees l from the forest F and r; join the created new tree to forest F. Finally, the delay of the AND item is set as the delay of the root node of the Huffman tree.

其中,步骤4包括:Among them, step 4 includes:

步骤41,计算最简RM表达式中含有的与项数m;Step 41, calculating the number m of AND terms contained in the simplest RM expression;

步骤42,为m个与项创建m个叶子结点;Step 42, create m leaf nodes for m AND items;

步骤43,根据m个与项的延时,构建具有m棵二叉树的森林F;Step 43, construct a forest F with m binary trees according to the delay of m AND items;

步骤44,重复执行以下操作直到森林F中仅具有一棵二叉树(即Huffman树):从森林F中选择根结点延时最小的两棵树l和r;创建一棵以l作为左子树并以r作为右子树的新树,其中,新树根结点的延时等于两子树延时中较大的加1;从森林F中删除被选择的两棵树l和r;将新创建的树加入森林F;Step 44, repeat the following operations until there is only one binary tree (i.e. Huffman tree) in the forest F: select the two trees l and r with the smallest root node delay from the forest F; create a tree with l as the left subtree And take r as the new tree of the right subtree, wherein, the delay of the root node of the new tree is equal to the larger of the delays of the two subtrees plus 1; delete the selected two trees l and r from the forest F; The newly created tree joins the forest F;

步骤45,将该最简RM表达式的延时设置为Huffman树根结点的延时。Step 45, setting the delay of the simplest RM expression as the delay of the root node of the Huffman tree.

本发明提供的一种单位延时模型下RM逻辑电路延时优化方法,通过对RM逻辑表达式进行化简使其与项数最少,可为后续获得具有最小延时的RM逻辑电路奠定基础;充分利用Huffman树构造算法可快速获得带权路径长度最小二叉树的优点,分别对最简RM逻辑表达式中的每个与项及最简RM逻辑表达式进行延时分解,可使每个与项及最简RM逻辑表达式的延时最小。本发明提供的一种单位延时模型下RM逻辑电路延时优化方法,可快速有效地将布尔逻辑电路转换为具有最小延时的RM逻辑电路,进而降低电路延时、提高电路的工 作速度和性能。The delay optimization method of RM logic circuit under the unit delay model provided by the present invention can lay the foundation for subsequent acquisition of RM logic circuits with minimum delay by simplifying the RM logic expression to minimize the number of terms; Making full use of the Huffman tree construction algorithm can quickly obtain the advantages of the minimum binary tree with weighted path length, and decompose each AND item in the simplest RM logic expression and the simplest RM logic expression separately, so that each AND item And the delay of the simplest RM logic expression is the smallest. The delay optimization method of RM logic circuit under a unit delay model provided by the present invention can quickly and effectively convert a Boolean logic circuit into an RM logic circuit with a minimum delay, thereby reducing circuit delay, improving circuit operating speed and performance.

【附图说明】【Description of drawings】

图1是本发明的一种单位延时模型下RM逻辑电路延时优化方法的流程图。FIG. 1 is a flow chart of an RM logic circuit delay optimization method under a unit delay model of the present invention.

图2是本发明的一种单位延时模型下RM逻辑电路延时优化方法一实施例的与项延时分解过程图,其中,结点上的数字代表以该结点为根结点的二叉树的延时。Fig. 2 is an AND term delay decomposition process diagram of an embodiment of an RM logic circuit delay optimization method under a unit delay model of the present invention, wherein the numbers on the nodes represent a binary tree with the node as the root node delay.

图2a是本发明实施例为5个输入变量创建5个叶子结点。Fig. 2a shows that the embodiment of the present invention creates 5 leaf nodes for 5 input variables.

图2b是本发明实施例创建一棵以两棵根结点延时最小的树为左右子树的新树。Fig. 2b shows the embodiment of the present invention to create a new tree with the two trees with the smallest root node delay as the left and right subtrees.

图2c是本发明实施例创建一棵以两棵根结点延时最小的树为左右子树的新树。Fig. 2c shows the embodiment of the present invention to create a new tree with the two trees with the smallest root node delay as the left and right subtrees.

图2d是本发明实施例创建一棵以两棵根结点延时最小的树为左右子树的新树。Fig. 2d is the embodiment of the present invention to create a new tree with the two trees with the smallest root node delay as the left and right subtrees.

图2e是本发明实施例形成一棵Huffman树。Fig. 2e is a Huffman tree formed by the embodiment of the present invention.

图3是本发明一种单位延时模型下RM逻辑电路延时优化方法一实施例的与项延时分解后的电路形式图。FIG. 3 is a circuit form diagram of an AND term delay decomposition of an embodiment of an RM logic circuit delay optimization method under a unit delay model of the present invention.

图4是本发明一种单位延时模型下RM逻辑电路延时优化方法一实施例的最简FPRM逻辑表达式的延时分解过程图。FIG. 4 is a delay decomposition process diagram of the simplest FPRM logic expression in an embodiment of an RM logic circuit delay optimization method under a unit delay model of the present invention.

【具体实施方式】【detailed description】

以下结合附图和具体实施例对本发明进行详细描述,但不作为对本发明的限定。The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments, but not as a limitation of the present invention.

图1是本发明的一种单位延时模型下RM逻辑电路延时优化方法的流程图。如图1所示,该方法包括:FIG. 1 is a flow chart of an RM logic circuit delay optimization method under a unit delay model of the present invention. As shown in Figure 1, the method includes:

步骤1,读入Boolean逻辑电路;Step 1, read in the Boolean logic circuit;

步骤2,利用RM表达式化简方法(如代数化简法、无关项化简法等)得到含与项数最少的最简RM逻辑表达式;Step 2, using the RM expression simplification method (such as algebraic simplification method, irrelevant item simplification method, etc.) to obtain the simplest RM logic expression containing the least number of terms;

步骤3,基于Huffman树构造算法对最简RM逻辑表达式中的每个与项进行延时分解,使得每个与项的延时最小;Step 3, based on the Huffman tree construction algorithm, decompose the delay of each AND item in the simplest RM logic expression, so that the delay of each AND item is the smallest;

步骤4,基于Huffman树构造算法对由所有与项组成的最简RM逻辑表达式 进行延时分解,使得最简RM逻辑表达式的延时最小;Step 4, based on the Huffman tree construction algorithm, perform delay decomposition on the simplest RM logic expression composed of all AND items, so that the delay of the simplest RM logic expression is the smallest;

步骤5,输出最简RM逻辑表达式的最小延时。Step 5, output the minimum delay of the simplest RM logic expression.

其中,步骤3包括:Among them, step 3 includes:

步骤31,计算与项中含有的输入变量数n;Step 31, calculating the number n of input variables contained in the AND item;

步骤32,若输入变量数n等于1,则将该与项的延时设置为该输入变量的延时;若输入变量数n大于1,则首先为n个输入变量创建n个叶子结点,其次根据d1,d2,...,di,...,dn(1≤i≤n,di代表第i个输入变量的延时)构建具有n棵二叉树的森林F={T1,T2,...,Ti,...Tn},1≤i≤n,其中每棵二叉树Ti仅具有一个延时为di的结点,最后重复执行以下操作直到森林F中仅具有一棵二叉树(即Huffman树):从森林F中选择根结点延时最小的两棵树l和r;创建一棵以l作为左子树并以r作为右子树的新树,其中,新树根结点的延时等于两子树延时中较大的加1;从森林F中删除被选择的两棵树l和r;将创建的新树加入森林F。最后将该与项的延时设置为Huffman树根结点的延时。Step 32, if the number of input variables n is equal to 1, set the delay of the AND item as the delay of the input variable; if the number of input variables n is greater than 1, first create n leaf nodes for n input variables, Secondly , construct a forest with n binary trees F= { T 1 ,T 2 ,...,T i ,...T n }, 1≤i≤n, where each binary tree T i has only one node with a delay of d i , and finally repeat the following operations until There is only one binary tree (ie Huffman tree) in the forest F: select two trees l and r with the smallest root node delay from the forest F; create a tree with l as the left subtree and r as the right subtree Create a new tree, where the delay of the root node of the new tree is equal to the greater of the delays of the two subtrees plus 1; delete the selected two trees l and r from the forest F; add the created new tree to the forest F. Finally, the delay of the AND item is set as the delay of the root node of the Huffman tree.

其中,步骤4包括:Among them, step 4 includes:

步骤41,计算最简RM表达式中含有的与项数m;Step 41, calculating the number m of AND terms contained in the simplest RM expression;

步骤42,为m个与项创建m个叶子结点;Step 42, create m leaf nodes for m AND items;

步骤43,根据m个与项的延时,构建具有m棵二叉树的森林F;Step 43, construct a forest F with m binary trees according to the delay of m AND items;

步骤44,重复执行以下操作直到森林F中仅具有一棵二叉树(即Huffman树):从森林F中选择根结点延时最小的两棵树l和r;创建一棵以l作为左子树并以r作为右子树的新树,其中,新树根结点的延时等于两子树延时中较大的加1;从森林F中删除被选择的两棵树l和r;将新创建的树加入森林F;Step 44, repeat the following operations until there is only one binary tree (i.e. Huffman tree) in the forest F: select the two trees l and r with the smallest root node delay from the forest F; create a tree with l as the left subtree And take r as the new tree of the right subtree, wherein, the delay of the root node of the new tree is equal to the larger of the delays of the two subtrees plus 1; delete the selected two trees l and r from the forest F; The newly created tree joins the forest F;

步骤45,将该最简RM表达式的延时设置为Huffman树根结点的延时。以下列举本发明的一种单位延时模型下RM逻辑电路延时优化方法一实施例。以一个5输入变量的布尔逻辑函数Step 45, setting the delay of the simplest RM expression as the delay of the root node of the Huffman tree. An embodiment of an RM logic circuit delay optimization method under a unit delay model of the present invention is listed below. Boolean logic function with a 5 input variable

为例,为得到具有最小延时的FPRM逻辑电路,第二参数X1代表第一个输入变量,X2代表第二个输入变量,X3代表第三个输入变量,X4代表第四个输入变量,X5代表第五个输入变量。该实施例的一种单位延时模型下RM逻辑电路延时优化方法包括: For example, in order to obtain the FPRM logic circuit with the minimum delay, the second parameter X 1 represents the first input variable, X 2 represents the second input variable, X 3 represents the third input variable, X 4 represents the fourth input variable, X 5 represents the fifth input variable. An RM logic circuit delay optimization method under a unit delay model of this embodiment includes:

步骤1,读入5变量布尔逻辑函数f(x5,x4,x3,x2,x1);Step 1, read in the 5-variable Boolean logic function f(x 5 ,x 4 ,x 3 ,x 2 ,x 1 );

步骤2,利用无关项化简法得到含与项数最少的最简FPRM(固定极性RM)逻辑表达式 Step 2, use the irrelevant item simplification method to obtain the simplest FPRM (fixed polarity RM) logic expression with the least number of AND items

步骤3,基于Huffman树构造算法对最简FPRM逻辑表达式中的每个与项进行延时分解,使得每个与项的延时最小。以与项为例,为简便起见,假定输入信号x5,x4,x3,x2,x1的延时都为零,其延时分解过程如图2所示,分解后的电路形式如图3所示:Step 3, based on the Huffman tree construction algorithm, decompose the delay of each AND term in the simplest FPRM logic expression, so that the delay of each AND term is the smallest. with the item For example, for the sake of simplicity, assume that the delays of the input signals x 5 , x 4 , x 3 , x 2 , and x 1 are all zero, and the delay decomposition process is shown in Figure 2, and the circuit form after decomposition is shown in Figure 3 Shown:

步骤31,计算与项中含有的输入变量数n,即n=5;Step 31, calculating the number of input variables n contained in the AND item, i.e. n=5;

步骤32,首先为5个输入变量创建5个叶子结点,其次根据5个输入变量的延时构建具有5棵二叉树的森林F,最后重复执行以下操作直到森林F中仅具有一棵二叉树(即Huffman树):从森林F中选择根结点延时最小的两棵树l和r;创建一棵以l作为左子树并以r作为右子树的新树,其中,新树根结点的延时等于两子树延时中较大的加1;从森林F中删除被选择的两棵树l和r;将新创建的树加入森林F。最后将该与项的延时设置为Huffman树根结点的延时,即为3。Step 32, first create 5 leaf nodes for 5 input variables, secondly construct a forest F with 5 binary trees according to the delay of 5 input variables, and finally repeat the following operations until there is only one binary tree in the forest F (ie Huffman tree): Select two trees l and r with the smallest root node delay from the forest F; create a new tree with l as the left subtree and r as the right subtree, where the new tree root node The delay is equal to the greater of the delays of the two subtrees plus 1; delete the selected two trees l and r from the forest F; add the newly created tree to the forest F. Finally, set the delay of the AND item to the delay of the root node of the Huffman tree, which is 3.

步骤4,基于Huffman树构造算法对由所有与项组成的最简FPRM逻辑表达式进行延时分解,使得最简FPRM逻辑表达式的延时最小,其延时分解过程如图4所示:Step 4: Based on the Huffman tree construction algorithm, the delay decomposition of the simplest FPRM logic expression composed of all AND items is carried out, so that the delay of the simplest FPRM logic expression is the smallest. The delay decomposition process is shown in Figure 4:

步骤41,计算最简FPRM表达式中含有的与项数m,即m=3;Step 41, calculate the number of AND terms m contained in the simplest FPRM expression, i.e. m=3;

步骤42,为3个与项创建3个叶子结点;Step 42, create 3 leaf nodes for 3 AND items;

步骤43,根据3个与项的延时,构建具有3棵二叉树的森林F;Step 43, construct a forest F with 3 binary trees according to the delay of 3 AND items;

步骤44,重复执行以下操作直到森林F中仅具有一棵二叉树(即Huffman 树):从森林F中选择根结点延时最小的两棵树l和r;创建一棵以l作为左子树并以r作为右子树的新树,其中,新树根结点的延时等于两子树延时中较大的加1;从森林F中删除被选择的两棵树l和r;将新创建的树加入森林F。Step 44, repeat the following operations until there is only one binary tree (i.e. Huffman tree) in the forest F: select two trees l and r with the smallest root node delay from the forest F; create a tree with l as the left subtree And take r as the new tree of the right subtree, wherein, the delay of the root node of the new tree is equal to the larger of the delays of the two subtrees plus 1; delete the selected two trees l and r from the forest F; The newly created tree joins the forest F.

步骤45,将该最简RM表达式的延时设置为Huffman树根结点的延时,即为4。Step 45, set the delay of the simplest RM expression as the delay of the root node of the Huffman tree, which is 4.

步骤5,输出最简FPRM逻辑表达式的最小延时。Step 5, output the minimum delay of the simplest FPRM logic expression.

当然,本发明还可有其它多种实施例,在不背离本发明精神及其实质的情况下,熟悉本领域的技术人员当可根据本发明作出各种相应的改变和变形,但这些相应的改变和变形都应属于本发明所附的权利要求的保护范围。Certainly, the present invention also can have other multiple embodiments, without departing from the spirit and essence of the present invention, those skilled in the art can make various corresponding changes and deformations according to the present invention, but these corresponding Changes and deformations should belong to the scope of protection of the appended claims of the present invention.

Claims (6)

1.一种单位延时模型下RM逻辑电路延时优化方法,其特征在于:该方法具体步骤包括:1. RM logic circuit delay optimization method under a unit delay model, it is characterized in that: the method concrete steps comprise: 步骤1,读入布尔Boolean逻辑电路;Step 1, read in Boolean logic circuit; 步骤2,利用RM表达式化简方法得到含与项数最少的最简RM逻辑表达式;Step 2, using the RM expression simplification method to obtain the simplest RM logic expression with the least number of AND items; 步骤3,基于哈夫曼Huffman树构造算法对最简RM逻辑表达式中的每个与项进行延时分解,使得每个与项的延时最小;Step 3, based on the Huffman Huffman tree construction algorithm, decompose the delay of each AND item in the simplest RM logic expression, so that the delay of each AND item is the smallest; 步骤4,基于Huffman树构造算法对由所有与项组成的最简RM逻辑表达式进行延时分解,使得最简RM逻辑表达式的延时最小;Step 4, based on the Huffman tree construction algorithm, perform delay decomposition on the simplest RM logic expression composed of all AND items, so that the delay of the simplest RM logic expression is the smallest; 步骤5,输出最简RM逻辑表达式的最小延时;Step 5, output the minimum delay of the simplest RM logic expression; 其中,RM含义为里德穆勒Reed-Muller。Among them, RM means Reed-Muller. 2.根据权利要求1所述的一种单位延时模型下RM逻辑电路延时优化方法,其特征在于:步骤2中的化简方法为代数化简法或无关项化简法。2. The delay optimization method for RM logic circuits under a unit delay model according to claim 1, characterized in that: the simplification method in step 2 is an algebraic simplification method or an irrelevant item simplification method. 3.根据权利要求1所述的一种单位延时模型下RM逻辑电路延时优化方法,其特征在于:步骤3包括:3. RM logic circuit delay optimization method under a kind of unit delay model according to claim 1, is characterized in that: step 3 comprises: 步骤31,计算与项中含有的输入变量数n;Step 31, calculating the number n of input variables contained in the AND item; 步骤32,若输入变量数n等于1,则将该与项的延时设置为该输入变量的延时;若输入变量数n大于1,则首先为n个输入变量创建n个叶子结点,其次根据d1,d2,...,di,...,dn构建具有n棵二叉树的森林F={T1,T2,...,Ti,...Tn},1≤i≤n,其中每棵二叉树Ti仅具有一个延时为di的结点,最后重复执行以下操作直到森林F中仅具有一棵二叉树:从森林F中选择根结点延时最小的两棵树l和r;创建一棵以l作为左子树并以r作为右子树的新树,1≤i≤n,di代表第i个输入变量的延时,dn代表第n个输入变量的延时。Step 32, if the number of input variables n is equal to 1, set the delay of the AND item as the delay of the input variable; if the number of input variables n is greater than 1, first create n leaf nodes for n input variables, Secondly, according to d 1 ,d 2 ,...,d i ,...,d n construct a forest F={T 1 ,T 2 ,...,T i ,...T n } with n binary trees , 1≤i≤n, where each binary tree T i has only one node with a delay of d i , and finally repeat the following operations until there is only one binary tree in the forest F: select the root node delay from the forest F The smallest two trees l and r; create a new tree with l as the left subtree and r as the right subtree, 1≤i≤n, d i represents the delay of the i-th input variable, d n represents Delay for the nth input variable. 4.根据权利要求1所述的一种单位延时模型下RM逻辑电路延时优化方法,其特征在于:步骤4包括:4. the RM logic circuit delay optimization method under a kind of unit delay model according to claim 1, is characterized in that: step 4 comprises: 步骤41,计算最简RM表达式中含有的与项数m;Step 41, calculating the number m of AND terms contained in the simplest RM expression; 步骤42,为m个与项创建m个叶子结点;Step 42, create m leaf nodes for m AND items; 步骤43,根据m个与项的延时,构建具有m棵二叉树的森林F;Step 43, construct a forest F with m binary trees according to the delay of m AND items; 步骤44,重复执行以下操作直到森林F中仅具有一棵二叉树,从森林F中选择根结点延时最小的两棵树l和r;创建一棵以l作为左子树并以r作为右子树的新树;Step 44, repeat the following operations until there is only one binary tree in the forest F, select two trees l and r with the smallest root node delay from the forest F; create a tree with l as the left subtree and r as the right the new tree of the subtree; 步骤45,将该最简RM表达式的延时设置为Huffman树根结点的延时。Step 45, setting the delay of the simplest RM expression as the delay of the root node of the Huffman tree. 5.根据权利要求3所述的一种单位延时模型下RM逻辑电路延时优化方法,其特征在于:新树根结点的延时等于两子树延时中较大的加1;从森林F中删除被选择的两棵树l和r;将创建的新树加入森林F;最后将该与项的延时设置为Huffman树根结点的延时。5. RM logic circuit time delay optimization method under a kind of unit time delay model according to claim 3, is characterized in that: the time delay of new tree root node is equal to bigger plus 1 in the time delay of two subtrees; From Delete the two selected trees l and r in the forest F; add the new tree created to the forest F; finally set the delay of the AND item to the delay of the root node of the Huffman tree. 6.根据权利要求4所述的一种单位延时模型下RM逻辑电路延时优化方法,其特征在于:新树根结点的延时等于两子树延时中较大的加1;从森林F中删除被选择的两棵树l和r;将新创建的树加入森林F。6. RM logic circuit time-delay optimization method under a kind of unit time-delay model according to claim 4, is characterized in that: the time delay of new tree root node is equal to bigger plus 1 in the time delay of two subtrees; From Delete the two selected trees l and r in forest F; add the newly created tree to forest F.
CN201610341295.8A 2016-05-20 2016-05-20 RM logic circuit delay optimization method in unit delay model Pending CN106027032A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610341295.8A CN106027032A (en) 2016-05-20 2016-05-20 RM logic circuit delay optimization method in unit delay model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610341295.8A CN106027032A (en) 2016-05-20 2016-05-20 RM logic circuit delay optimization method in unit delay model

Publications (1)

Publication Number Publication Date
CN106027032A true CN106027032A (en) 2016-10-12

Family

ID=57096720

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610341295.8A Pending CN106027032A (en) 2016-05-20 2016-05-20 RM logic circuit delay optimization method in unit delay model

Country Status (1)

Country Link
CN (1) CN106027032A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107194023A (en) * 2017-03-30 2017-09-22 宁波大学 A kind of FPRM circuit areas and delay Optimization method
CN109710973A (en) * 2018-11-22 2019-05-03 温州大学 A three-valued fixed-polarity RM circuit area, power consumption and delay optimization method

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102054102A (en) * 2010-12-27 2011-05-11 宁波大学 Best mixed polarity searching method of AND/XOR circuit
CN102592013A (en) * 2011-12-31 2012-07-18 宁波大学 Optimization method for time delay and area of fixed-polarity Reed-Muller circuit

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102054102A (en) * 2010-12-27 2011-05-11 宁波大学 Best mixed polarity searching method of AND/XOR circuit
CN102592013A (en) * 2011-12-31 2012-07-18 宁波大学 Optimization method for time delay and area of fixed-polarity Reed-Muller circuit

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
LIMIN XIAO: "Optimization of Best Polarity Searching for Mixed Polarity Reed-Muller Logic Circuit", 《SYSTEM-ON-CHIP CONFERENCE (SOCC), 2015 28TH IEEE INTERNATIONAL》 *
王振海: "Reed_Muller逻辑电路的延时优化", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107194023A (en) * 2017-03-30 2017-09-22 宁波大学 A kind of FPRM circuit areas and delay Optimization method
CN107194023B (en) * 2017-03-30 2019-07-12 宁波大学 An FPRM circuit area and delay optimization method
CN109710973A (en) * 2018-11-22 2019-05-03 温州大学 A three-valued fixed-polarity RM circuit area, power consumption and delay optimization method

Similar Documents

Publication Publication Date Title
CN102054102A (en) Best mixed polarity searching method of AND/XOR circuit
Sinha et al. Design and analysis of low power 1-bit full adder cell
Adilakshmi Siliveru Design of Kogge-stone and brent-kung adders using Degenerate pass Transistor logic
CN105187051A (en) Power and area optimization method of incomplete certain Reed-Muller circuit based on NSGA-II
CN102982205B (en) A kind of fixed polarity conversion method for Design of Digital Circuit
CN104881549B (en) A kind of power consumption optimization method of the Reed Muller logic circuits comprising outlier
Zhao et al. Efficient ternary logic circuits optimized by ternary arithmetic algorithms
Shrivas et al. Design and performance analysis of 1 bit full adder using GDI technique in nanometer era
CN106027032A (en) RM logic circuit delay optimization method in unit delay model
Jayanthi et al. Comparison of performance of high speed VLSI adders
Chu et al. A high-performance design of generalized pipeline cellular array
CN105334906A (en) Multistage gated clock network optimization method in nanometer technology
Wang et al. Power optimization for FPRM logic using approximate computing technique
Jaekel et al. Design of dynamic pass-transistor logic circuits using 123 decision diagrams
Saji et al. A low power variable sized CSLA implementation using GDI logic in 45nm SOI technology
CN109032561B (en) Reversible logic adder circuit with carry bypass output as carry selection
Devi et al. An asynchronous low power and high performance VLSI architecture for Viterbi decoder implemented with quasi delay insensitive templates
Lueangsongchai et al. Design high speed and low power hybrid full adder circuit
Thakur et al. CMOS design of area and power efficient multiplexer using tree topology
Samundiswary et al. Design and Analysis of CMOS Based DADDA Multiplier
Vijayakumari et al. An improved design of combinational digital circuits with multiplexers using genetic algorithm
CN202617076U (en) Early termination asynchronous comparator
Dai et al. An Efficient Spiking Convolutional Architecture with Compressed Address Event Representation and Adaptive Delay Asynchronous Clocks
CN103885748B (en) A kind of low-power consumption random number post-processing approach
Ma et al. Power optimization based on dual-logic using And-Xor-Inverter Graph

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20161012