The present application claims the priority to Chinese Patent Application No. 201810501722.3, titled “SIGNAL INTEGRITY SIMULATION METHOD FOR ENCRYPTION HYBRID MODEL”, filed on May 23, 2018 with the Chinese Patent Office, which is incorporated herein by reference in its entirety.
The present disclosure relates to the technical field of high-speed signal quality evaluation, and in particular to a signal integrity simulation method for an encryption hybrid model.
With the continuous development of computer technologies, a high speed interconnection phenomenon is playing a leading role in many factors that reflect the performance of digital systems. In the previous high-speed system designs, a method of “wiring-simulation-rewiring” is often adopted, which requires a design engineer to perform wiring and then perform simulation operations. If an error is found after performing the wiring, it is required to find the cause and modify the design, and then perform the above process of “wiring-simulation-rewiring” again.
In order to meet a design requirement, much time may be taken with this design. Therefore, a design engineer has to use an appropriate method and analysis simulation tool in early design stage to assess the feasibility and risk points of the system design in advance, and optimize the design based on the simulation result, to improve success rate of the system design and shorten a development cycle.
However, some chip manufacturers encrypt their chip models, and there are also some cases that an actual type a chip model is not consistent with a given type of the chip model, which may increase the simulation difficulty and reduce the simulation accuracy in a simulation process of the early design stage, and even make the simulation cannot be performed normally. In this case, it may be impossible to effectively evaluate link characteristics, which may increase risks in the design.
In the conventional technical solution, a universal chip model is used to approximately replace the chip model in the link, to approximately assess the link characteristics of the system design. Although this approximation can realize link simulation, it cannot guarantee the accuracy of the simulation. Especially when the link margin is very small, this simulation method is not very meaningful.
Therefore, it is required to provide a signal integrity simulation method for an encryption hybrid model.
A signal integrity simulation method for an encryption hybrid model is provided according to the present disclosure, to solve the problem that the simulation result is distorted or the link simulation cannot be performed in the signal integrity simulation in the early design stage of a high speed circuit with a relay chip due to that the adopted chip model is encrypted or has a type inconsistent with that of other models.
The technical solutions adopted by the present disclosure are described as follows.
A signal integrity simulation method for an encryption hybrid model is provided according to the embodiments of the present disclosure, which includes:
extracting a step response data of a SPICE model through a transient simulation;
generating an external random code signal;
importing the extracted step response data as an input source for an ADS channel simulator; and
calculating, by the ADS channel simulator, an eye diagram at a RX end based on an algorithm and the external random code signal, to perform measurement on the eye diagram.
Further, a link verification is formed before performing the transient simulation, and the link verification is formed by connecting a TX end to the RX end via a wiring model. A chip at the TX end is a SPICE model that is encrypted, and a step response generated by the chip at the TX end includes all characteristics of the chip at the TX end. A chip at the RX end is an IBIS AMI model. The SPICE model at the TX end is a RC circuit.
Further, in a case that the SPICE model at the TX end is not encrypted, the simulation method includes the following steps 1) to 5). In step 1), the SPICE model is imported based on a Spice Wizard function of ADS. The SPICE model at the TX end that is not encrypted is capable of being directly called by the ADS. An input interface is reserved for the SPICE model, to input a step signal outside the SPICE model. In step 2), a link model is established. In step 3), a transient step response simulation is performed. The imported SPICE model is connected to a wiring model, and a resistor of 50 ohm is joined at an end of the verification link. An ideal step signal is inputted at the input interface reserved for the SPICE model, a simulation time is set to make an output of the step response be in a steady state, and a simulation time interval is set to guarantee calculation accuracy. The step response data is extracted at a position where the resistor is joined, where the step response data includes the characteristics of the chip at the TX end and the verification link. In step 4), a channel simulation is performed. The step response data obtained in step 3) is imported into the channel simulation as a signal source at the RX end, and the external random code signal is inputted, and parameters related to the channel simulation are set. In step 5), simulation results are verified. A channel simulation result obtained in step 4) is compared with a transient simulation result, to verify consistency between the channel simulation result and the transient simulation result.
Further, in a case that the SPICE model at the TX end is encrypted, the simulation method includes the following steps 1) to 3). In step 1), a transient simulation based on HSPICE is performed. A transient step response simulation is performed on the encrypted SPICE model and the verification link to obtain a step response output file with a suffix of .lis. A format of the step response output file is edited to form a file with a suffix of .tim that is capable of being called by ADS. In step 2), An active simulation is performed. The step response data (.tim) is imported into the ADS for the active simulation, and the active simulation is performed by calling the step response data in the ADS based on the external random code signal. In step 3), simulation results are verified. A transient simulation result is compared with an active simulation result obtained in step 2) to verify consistency between the transient simulation result and the active simulation result.
Compared with the conventional technology, the advantageous effects of the embodiments of the present disclosure are as follows. With the signal integrity simulation method for an encryption hybrid model such as an IBIS AMI model according to the present disclosure, the problem that signal integration simulation cannot be accurately performed or cannot be performed because models provided by the manufacturer are encrypted or the types of models provided by the manufacturer are not consistent can be effectively solved. With the technical solution of the present disclosure, it is ensured that link quality analysis can be performed according to different types of encryption models. The system design risk can be assessed based on a simulation result and a corresponding improvement can be made, unnecessary time and expense cost are reduced, and design success rate of the system is greatly improved.
In addition, the method according to the present disclosure has a reliable principle and simple steps, thus has a wide application prospect.
It can be seen from the above that, compared with the conventional technology, the technical solution of the present disclosure has an outstanding substantive feature and a significant progress, and the beneficial effect of the implementation thereof is also obvious.
The embodiments of the present disclosure are described in conjunction with drawings.
In the embodiment, a signal integrity simulation method for an encryption hybrid model is provided, which includes:
extracting a step response data of a SPICE model through a transient simulation;
generating an external random code signal;
importing the extracted step response data as an input source of an ADS channel simulator; and
calculating, by the ADS channel simulator, an eye diagram at a RX end based on an algorithm and the external random code signal, to perform measurement on the eye diagram.
In the embodiment, a link verification is formed before performing the transient simulation, where the link verification is formed by connecting a TX end to the RX end via a wiring model. A chip at the TX end is a SPICE model that is encrypted, and a step response generated by the chip at the TX end includes all characteristics of the chip at the TX end. A chip at the RX end is an IBIS AMI model. The SPICE model at the TX end is a RC circuit.
In the embodiment, the SPICE model at the TX end is not encrypted. In this case, the simulation method includes the following steps 1) to 5).
In step 1), the SPICE model is imported based on a Spice Wizard function of ADS. The SPICE model at the TX end that is not encrypted is capable of being directly called by the ADS. An input interface is reserved for the SPICE model, to input a step signal outside the SPICE model.
In step 2), a link model is established.
In step 3), a transient step response simulation is performed. The imported SPICE model is connected to a wiring model, and a resistor of 50 ohm is joined at an end of the verification link. An ideal step signal is inputted at the input interface reserved for the SPICE model, a simulation time is set to make an output of the step response be in a steady state, and a simulation time interval is set to guarantee calculation accuracy. The step response data is extracted at a position where the resistor is joined, where the step response data includes the characteristics of the chip at the TX end and the verification link.
In step 4), a channel simulation is performed. The step response data obtained in step 3) is imported into the channel simulation as a signal source at the RX end, and the external random code signal is inputted, and parameters related to the channel simulation are set.
In step 5), simulation results are verified. A channel simulation result obtained in step 4) is compared with a transient simulation result, to verify consistency between the channel simulation result and the transient simulation result.
In the embodiment, a signal integrity simulation method for an encryption hybrid model is provided. In the embodiment, the SPICE model at the TX end is encrypted. In this case, the simulation method includes the following steps 1) to 3).
In step 1), a transient simulation based on HSPICE is performed. A transient step response simulation is performed on the encrypted SPICE model and the verification link to obtain a step response output file with a suffix of .lis. A format of the step response output file is edited to form a file with a suffix of .tim that is capable of being called by ADS.
In step 2), An active simulation is performed. The step response data (.tim) is imported into the ADS for the active simulation, and the active simulation is performed by calling the step response data in the ADS based on the external random code signal. An input interface is reserved for the SPICE model, to input a step signal outside the SPICE model. A signal resource at the RX end is imported into the channel simulation, and the external random code signal is inputted, and parameters related to the active simulation are set.
In step 3), simulation results are verified. A transient simulation result is compared with an active simulation result obtained in step 2) to verify consistency between the transient simulation result and the active simulation result.
In the embodiment, a simulation link is as shown in
Since the SPICE model and the IBIS AMI model cannot exist in the simulation link together, the verification link shown in
In the embodiment, assuming that the SPICE model at the TX end is not encrypted, the simulation method includes the following steps 1) to 5).
In step 1), the SPICE model is imported. The SPICE model is imported based on a Spice Wizard function of ADS. In the embodiment, since the SPICE model at the TX end is not encrypted, the SPICE model may be called directly by the ADS. This step cannot be performed if the SPICE model is encrypted. An input interface is reserved for the SPICE model, to input a step signal outside the SPICE model.
In step 2), a link model is established. In the embodiment, it is only required to establish a wiring model of 5 inch length. Because of different extraction manners of the wiring model in actual projects, the wiring model may be transformed into an S parameter for the ADS to call in other embodiments of the present disclosure. In the embodiment, Pre_channel and Post_channel are both wiring models of 5 inch length.
In step 3), a transient step response simulation is performed. The imported SPICE model is connected to the wiring model of 5 inch length, and a resistor of 50 ohm is joined at an end of a link. An ideal step signal is inputted at the input interface reserved for the SPICE model, and an amplitude of the ideal step signal is determined depending on the chip characteristics. A simulation time is set to make an output of the step response be in a steady state, and a simulation time interval is set to guarantee calculation accuracy. The step response data is extracted at a position where the resistor is joined, and the step response data includes all the characteristics of the chip at the TX end and the link.
In step 4), a channel simulation is performed. The above step response data is imported into the channel simulation as a signal source of the chip at the RX end, and an external random code signal is inputted. The step response data includes the characteristics of the chip at the TX end and the link, and the external random code signal is inputted to generate an eye diagram. Relevant parameters, such as a simulation frequency, a code source type, equalization, are set for channel simulation, and the parameters are set according to actual situations. In the embodiment, a PCIE Gen3 signal is used, thus the frequency is set to be 8G.
In step 5), the simulation results are verified. A channel simulation result obtained in step 4) is compared with a transient simulation result, to verify the feasibility and accuracy of the method.
The simulation result based on the transient simulation method is as shown in
A result based on the above hybrid model simulation method is as shown in
Compared among the above simulation results, it can be seen that the simulation result based on the step response is almost consistent with the transient simulation result, which verifies the effectiveness of performing channel simulation by using the step response of the chip.
In the embodiment, assuming that SPICE model at the TX end is encrypted, the simulation method includes the following steps 1) to 2).
In step 1), a transient simulation based on HSPICE is performed. A transient step response simulation is performed on the encrypted SPICE model and the verification link (a wire of 5 inches) to obtain a step response output file with a suffix of .lis. A format of the step response output file is edited to form a file with a suffix of .tim that is capable of being called by the ADS.
In step 2), the step response data (.tim) is imported into the ADS for an active simulation. The active simulation is performed by calling the step response data in the ADS based on the external random code signal.
The simulation result obtained with the above simulation method are as shown in
The above are only preferred embodiments of the present disclosure. Those skilled in the art, based on the application method and principle of the present disclosure, can easily make possible variations and modifications to the technical solutions of the present disclosure, not limited to the methods described in the above embodiments of the present disclosure. Therefore, the methods described above is only preferred, and the present disclosure is not limited to the embodiments described herein.
Number | Date | Country | Kind |
---|---|---|---|
201810501722.3 | May 2018 | CN | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/CN2018/112050 | 10/26/2018 | WO | 00 |