CN112765918B - Fuzzy reasoning method for determining design parameters of integrated circuit - Google Patents

Abstract

The invention discloses a fuzzy reasoning method for determining design parameters of an integrated circuit, which comprises the following steps: building an integrated circuit design database; setting target performance indexes of the design parameters of the integrated circuit, subtracting the target performance indexes from each performance index in a database, and obtaining a plurality of performance index error vectors corresponding to each design parameter; calculating the membership degree of the performance index error vector by using a fuzzy reasoning method; and multiplying the membership degree of the performance index error vector with the corresponding design parameters in the database and summing to obtain the integrated circuit design parameters reaching the target performance index. The invention establishes a database according to the design data of the integrated circuit aiming at the difficult problem that the size parameters of the integrated circuit device are difficult to be determined quickly in the research and development process of the integrated circuit chip, rapidly estimates the design parameters reaching the target performance index by using a fuzzy reasoning method, can be applied to the design of the integrated circuit, and provides a new method for shortening the design period of the integrated circuit.

Description

A Fuzzy Reasoning Method for Determining Design Parameters of Integrated Circuits

technical field

The invention belongs to the field of integrated circuit design, and in particular relates to a fuzzy reasoning method for determining integrated circuit design parameters.

Background technique

The development of integrated circuits, especially analog integrated circuits, relies more on the design experience of R&D personnel and repeated iterative simulation verification of software. Designers constantly adjust the size parameters of devices in integrated circuits to obtain the desired performance indicators of circuit modules. This time-consuming and labor-intensive design method reduces the efficiency of integrated circuit design, increases the cost of research and development, and prolongs the development cycle of integrated circuit chips.

Contents of the invention

Aiming at the above-mentioned deficiencies in the prior art, a fuzzy reasoning method for determining integrated circuit design parameters provided by the present invention solves the problems existing in the prior art.

In order to achieve the above-mentioned purpose of the invention, the technical solution adopted in the present invention is: a fuzzy reasoning method for determining integrated circuit design parameters, comprising the following steps:

S1. Build an integrated circuit design database;

S2. Set the target performance index of the integrated circuit design parameter, and subtract it from each performance index in the database to obtain several performance index error vectors corresponding to each design parameter;

S3. Calculate the degree of membership of the error vector of the performance index by using the fuzzy reasoning method;

S4. Multiply and sum the membership degree of the error vector of the performance index and the corresponding design parameter in the database to obtain the design parameter of the integrated circuit that reaches the target performance index.

Further, the database in the step S1 includes several integrated circuit design case data, and the integrated circuit design case data includes design parameters and performance indicators.

Further, the step S2 includes the following steps:

S2.1. Setting target performance indicators of integrated circuit design parameters;

S2.2. Normalize both the target performance index and the performance index in the database;

S2.3. Subtracting the target performance index from each performance index in the database to obtain several performance index error vectors corresponding to each parameter.

Further, the function of the normalization process in the step S2 is:

Among them, x _n represents the normalized data, x represents the data that needs to be normalized, x _min represents the minimum value among all the data that needs to be normalized, and x _max represents the maximum value among all the data that needs to be normalized.

Further, the performance index error vector includes gain error, gain bandwidth product error and output noise error.

Further, the method for obtaining the degree of membership in the step S3 is specifically:

S3.1. Acquiring the initial degree of membership p of each performance index error in the performance index error vector is:

Among them, p represents the degree of membership of the performance index error vector, A represents the first intermediate parameter, e represents the normalized performance index error vector, and c represents the second intermediate parameter;

S3.2. Multiply all the initial membership degrees p of the performance index error vector to obtain the initial membership degree p _v of the performance index error vector as:

Among them, p _v represents the initial membership degree of the performance index error vector, p _i represents the membership degree corresponding to the i-th performance index error in the performance index error vector, i=1,2,...,n, n represents the performance index The total number of performance index errors in the error vector;

S3.3. Normalize the initial membership degree _pv of each performance index error vector, and obtain the membership degree of the performance index error vector as follows:

Among them, p _j represents the initial membership degree before the jth performance index error vector is normalized, p _jn represents the membership degree after the jth performance index error vector is normalized, j=1,2,...,m , m represents the total number of performance indicator error vectors.

Further, the step S4 is specifically:

S4.1. Express the degree of membership and the design parameters in the database as vectors P and D respectively;

S4.2. Multiply and sum the membership degree of the performance index error vector with the corresponding design parameters in the database, and obtain the integrated circuit design parameter Q that reaches the target performance index, specifically:

Q＝P D ^T

Among them, T means transpose.

The beneficial effects of the present invention are:

(1) The present invention aims at the difficult problem that the dimensional parameters of integrated circuit devices are difficult to quickly determine in the process of developing integrated circuit chips. A database is established according to the integrated circuit design data, and the design parameters that reach the target performance index are quickly estimated by using the fuzzy reasoning method, which can be applied In the design of integrated circuits, it provides a new method for shortening the design cycle of integrated circuits.

(2) The present invention has less calculation amount and low complexity, and can quickly determine the design parameters of the integrated circuit, which has important engineering significance for reducing the research and development cost of the integrated circuit chip.

Description of drawings

FIG. 1 is a flowchart of a fuzzy reasoning method for determining integrated circuit design parameters proposed by the present invention.

FIG. 2 is an integrated circuit diagram of an NMOS input differential pair with a resistive load applied in the implementation of the present invention.

Fig. 3 is an integrated circuit diagram of an NMOS input differential pair after the design parameters are acquired by the method of the present invention.

Figure 4 shows the results of simulation experiments using Cadence software based on the obtained IC design parameters: gain and gain-bandwidth product.

Figure 5 shows the results of simulation experiments using Cadence software based on the obtained IC design parameters: output noise.

Detailed ways

Specific embodiments of the present invention are described below so that those skilled in the art can understand the present invention, but it should be clear that the present invention is not limited to the scope of specific embodiments. For those of ordinary skill in the art, as long as various changes Within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions and creations using the concept of the present invention are included in the protection list.

Embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings.

As shown in Figure 1, a fuzzy reasoning method for determining integrated circuit design parameters includes the following steps:

S1. Build an integrated circuit design database;

S2. Set the target performance index of the integrated circuit design parameter, and subtract it from each performance index in the database to obtain several performance index error vectors corresponding to each design parameter;

S3. Calculate the degree of membership of the error vector of the performance index by using the fuzzy reasoning method;

S4. Multiply and sum the membership degree of the error vector of the performance index and the corresponding design parameter in the database to obtain the design parameter of the integrated circuit that reaches the target performance index.

The database in step S1 includes a number of integrated circuit design case data, and the integrated circuit design case data includes design parameters and performance indicators.

As shown in FIG. 2 , in this embodiment, parameters are determined for the resistive load of the NMOS input differential pair circuit.

Described step S2 comprises the following steps:

S2.1. Setting target performance indicators of integrated circuit design parameters;

S2.2. Normalize both the target performance index and the performance index in the database;

S2.3. Subtracting the target performance index from each performance index in the database to obtain several performance index error vectors corresponding to each parameter.

The function of the normalization process in the step S2 is:

Among them, x _n represents the normalized data, x represents the data that needs to be normalized, x _min represents the minimum value among all the data that needs to be normalized, and x _max represents the maximum value among all the data that needs to be normalized.

The performance index error vector includes gain error, gain bandwidth product error and output noise error.

In this embodiment, the performance index error vector is the difference between the target performance index and the performance index stored in the database, which can be expressed as:

E _i =[G _des -G _i ,GBW _des -GBW _i ,ON _des -ON _i ]i=1,2,...m

Among them, G _des , GBW _des and ON _des represent the target performance indicators (gain, gain-bandwidth product and output noise) respectively, and G _i , GBW _i and ON _i represent the performance indicators (gain, gain-bandwidth product and output noise) stored in the database respectively. noise), E _i is the i-th performance index error vector, and m is the number of performance index error vectors.

The method for obtaining the degree of membership in the step S3 is specifically:

S3.1. Acquiring the initial degree of membership p of each performance index error in the performance index error vector is:

Among them, p represents the degree of membership of the performance index error vector, A represents the first intermediate parameter, e represents the normalized performance index error vector, and c represents the second intermediate parameter;

S3.2. Multiply all the initial membership degrees p of the performance index error vector to obtain the initial membership degree p _v of the performance index error vector as:

Among them, p _v represents the initial membership degree of the performance index error vector, p _i represents the membership degree corresponding to the i-th performance index error in the performance index error vector, i=1,2,...,n, n represents the performance index The total number of performance index errors in the error vector;

S3.3. Normalize the initial membership degree _pv of each performance index error vector, and obtain the membership degree of the performance index error vector as follows:

Among them, p _j represents the initial membership degree before the jth performance index error vector is normalized, p _jn represents the membership degree after the jth performance index error vector is normalized, j=1,2,...,m , m represents the total number of performance indicator error vectors.

The step S4 is specifically:

S4.1. Express the degree of membership and the design parameters in the database as vectors P and D respectively;

S4.2. Multiply and sum the membership degree of the performance index error vector with the corresponding design parameters in the database, and obtain the integrated circuit design parameter Q that reaches the target performance index, specifically:

Q＝P D ^T

Among them, T means transpose.

In this embodiment, the degree of membership of the performance indicator error vector and the design parameters in the database are expressed in vector form:

P=[p _1n ,p _2n ,...,p _mn ], I=[i ₁ ,i ₂ ,...,i _m ], C=[c ₁ ,c ₂ ,...,c _m ],

R＝[r ₁ ,r ₂ ,...,r _m ]

Among them, P is a vector composed of the membership degree of the performance index error vector, I, C and R represent the NMOS input tube size, the tail current source tube size and the load resistance value vector respectively, and the estimated design parameters to achieve the target performance index can be expressed as :

i _o = P · I ^T

c _o ＝P·C ^T

r _o ＝P·R ^T

Among them, P is the membership degree of the error vector of the performance index, and i _o , c _o and r _o represent the size of the NMOS input tube, the size of the tail current source tube and the value of the load resistance obtained by fuzzy reasoning respectively.

According to the expected performance index (gain, gain-bandwidth product and output noise are 20dB, 144MHz and The design parameters of the NMOS input differential pair circuit with resistive load are quickly estimated by using the fuzzy reasoning method of the above-mentioned integrated circuit design parameters. The optimized input tube size, tail current source tube size (channel width/channel length) and load resistance values are 32μm/0.18μm, 18μm/0.18μm and 10KΩ, respectively, as shown in Figure 3. According to the estimated design parameters of the NMOS input differential pair circuit with resistive load, the simulation experiment results using Cadence software are shown in Figure 4-5. The gain, gain-bandwidth product, and output noise of the NMOS input differential pair circuit with resistive load are 20.78dB. , 140.00MHz and The expected performance index is achieved, which shows that the proposed method can effectively estimate the design parameters of the NMOS input differential pair circuit with resistive load, and obtain an integrated circuit that meets the performance requirements.

From the above results, it can be found that the method proposed by the present invention can simply and effectively estimate the design parameters of the integrated circuit, shorten the development cycle of the integrated circuit, and provide a reliable way for realizing the efficient design of the integrated circuit.

Claims (3)

1. A fuzzy reasoning method for determining integrated circuit design parameters is characterized in that, comprising the following steps: S1. Build an integrated circuit design database; S2. Set the target performance index of the integrated circuit design parameter, and subtract it from each performance index in the database to obtain several performance index error vectors corresponding to each design parameter; S3. Calculate the degree of membership of the error vector of the performance index by using the fuzzy reasoning method; S4. Multiply and sum the membership degree of the performance index error vector with the corresponding design parameters in the database to obtain the integrated circuit design parameters that meet the target performance index; Described step S2 comprises the following steps: S2.1. Setting target performance indicators of integrated circuit design parameters; S2.2. Normalize both the target performance index and the performance index in the database; S2.3. Subtracting the target performance index from each performance index in the database to obtain several performance index error vectors corresponding to each parameter; The function of the normalization process in the step S2 is: Among them, x _n represents the normalized data, x represents the data that needs to be normalized, x _min represents the minimum value of all data that needs to be normalized, and x _max represents the maximum value of all data that needs to be normalized; The performance index error vector includes gain error, gain bandwidth product error and output noise error; The method for obtaining the degree of membership in the step S3 is specifically: S3.1. Acquiring the initial degree of membership p of each performance index error in the performance index error vector is: Among them, p represents the degree of membership of the performance index error vector, A represents the first intermediate parameter, e represents the normalized performance index error vector, and c represents the second intermediate parameter; S3.2. Multiply all the initial membership degrees p of the performance index error vector to obtain the initial membership degree p _v of the performance index error vector as: Among them, p _v represents the initial membership degree of the performance index error vector, p _i represents the membership degree corresponding to the i-th performance index error in the performance index error vector, i=1,2,...,n, n represents the performance index The total number of performance index errors in the error vector; S3.3. Normalize the initial membership degree _pv of each performance index error vector, and obtain the membership degree of the performance index error vector as follows: Among them, p _j represents the initial membership degree before the jth performance index error vector is normalized, p _jn represents the membership degree after the jth performance index error vector is normalized, j=1,2,...,m , m represents the total number of performance indicator error vectors. 2. The fuzzy reasoning method for determining integrated circuit design parameters according to claim 1, characterized in that, in the step S1, the database includes some integrated circuit design case data, and the integrated circuit design case data includes design parameters and performance indicators . 3. The fuzzy reasoning method for determining integrated circuit design parameters according to claim 1, wherein said step S4 is specifically: S4.1. Express the degree of membership and the design parameters in the database as vectors P and D respectively; S4.2. Multiply and sum the membership degree of the performance index error vector with the corresponding design parameters in the database, and obtain the integrated circuit design parameter Q that reaches the target performance index, specifically: Q＝P D ^T Among them, T means transpose.