CN111860774B - True random number-based eigen state network circuit signal preparation system and method - Google Patents
True random number-based eigen state network circuit signal preparation system and method Download PDFInfo
- Publication number
- CN111860774B CN111860774B CN202010618163.1A CN202010618163A CN111860774B CN 111860774 B CN111860774 B CN 111860774B CN 202010618163 A CN202010618163 A CN 202010618163A CN 111860774 B CN111860774 B CN 111860774B
- Authority
- CN
- China
- Prior art keywords
- random number
- neural network
- hopfield neural
- true random
- eigen
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F7/00—Methods or arrangements for processing data by operating upon the order or content of the data handled
- G06F7/58—Random or pseudo-random number generators
- G06F7/588—Random number generators, i.e. based on natural stochastic processes
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention relates to a preparation system of an eigen state network circuit signal based on a true random number, which comprises the following steps: the system comprises a random number generator, an SDN controller, a Hopfield neural network device and an Internet of things gateway. The Hopfield neural network device includes a plurality of discrete Hopfield neural network circuits. An SDN controller in communication with the random number generator and the Hopfield neural network device through the Internet of things gateway; the Hopfield neural network device selects the discrete Hopfield neural network circuit to generate the eigen-state network electrical signal based on the control strategy and the true random number stream. The invention also relates to a preparation method of the eigen state network circuit signal based on the true random number. The invention can rapidly prepare a large amount of network circuit signals with high randomness. By using three sets of physical random signals and generating random numbers only when at least two sets of signals are normal, it is ensured that a high quality circuit signal is produced.
Description
Technical Field
The invention relates to the technical field of quantum computation, in particular to a system and a method for preparing an eigen state network circuit signal based on a true random number.
Background
Both fault diagnosis, lifetime prediction, maintenance decisions related to major equipment facilities, accurate medical, biomedical, credit surveillance related to population health care, and personal privacy protection, infrastructure defense, national security related to cyberspace security require large-scale data analysis and rely on very computationally intensive load-bearing.
Under the condition that the demand of large-scale data analysis for super-high computing power is continuously increased, a network circuit signal with high randomness needs to be rapidly prepared in a large quantity. And there is no such method or system known in the art.
Disclosure of Invention
The present invention is directed to provide a system and a method for preparing an eigen-state network circuit signal based on a true random number, which can prepare a large amount of network circuit signals with high randomness quickly.
The technical scheme adopted by the invention for solving the technical problems is as follows: a preparation system of an eigen state network circuit signal based on a true random number is constructed, and comprises the following steps:
a random number generator for generating a stream of true random numbers based on the three sets of physical random signals;
an SDN controller for generating a control strategy based on the true random number flow and a preset;
a Hopfield neural network device comprising a plurality of discrete Hopfield neural network circuits;
an instrumented gateway through which the SDN controller communicates with the random number generator and the Hopfield neural network device;
wherein the Hopfield neural network device selects the discrete Hopfield neural network circuit to generate the eigenstate network electrical signal based on the control strategy and the stream of true random numbers.
In the eigen-state network circuit signal preparation system based on the true random number, the control strategy comprises that each discrete Hopfield neural network circuit corresponds to one true random number, and each discrete Hopfield neural network circuit receives one true random number and outputs one eigen-state network electric signal.
In the preparation system of the eigen-state network circuit signal based on the true random number, the discrete Hopfield neural network circuit judges whether the received true random number is larger than a set threshold value, if so, the discrete Hopfield neural network circuit outputs an eigen-state network electric signal 1, otherwise, the discrete Hopfield neural network circuit outputs an eigen-state network electric signal 0.
In the eigen-state network circuit signal preparation system based on the true random number, the set threshold is a fixed threshold or a dynamic threshold generated based on a random function.
In the eigen-state network circuit signal preparation system based on the true random number, the control strategy comprises that each discrete Hopfield neural network circuit corresponds to one true random number, each discrete Hopfield neural network circuit receives one true random number, and a plurality of discrete Hopfield neural network circuits output one eigen-state network electric signal.
In the eigen-state network circuit signal preparation system based on the true random number, the discrete Hopfield neural network circuits output eigen-state network electric signals 0 or eigen-state network electric signals 1 based on the received true random numbers and the set threshold value.
In the eigenstate network circuit signal preparation system based on the true random number, the Hopfield neural network device comprises n discrete Hopfield neural network circuits, each discrete Hopfield neural network circuit is used for generating a circuit signal represented by an n-dimensional vector Y (t) at the time t for the one-bit true random number received by the discrete Hopfield neural network circuit, wherein Y (t) ═[ Y (t) } is used for generating a circuit signal represented by an n-dimensional vector Y (t) at the time t1(t),Y2(t),...,Yn(t)]TN is an integer greater than zero, Yj(t) (j ═ 1 … … n) takes the value 1 or 0.
In the system for preparing an eigen-state network circuit signal based on a true random number, the random number generator further includes:
the light source random signal generating module is used for generating three groups of independent physical random signals;
a true random number generation module for generating a true random number stream based on the three sets of physical random signals;
and the verification module is used for verifying the safety and the randomness of the true random number by adopting a random statistical verification package.
In the eigen-state network circuit signal preparation system based on the true random number, the three groups of independent physical random signals comprise a light intensity signal, an electromagnetic radiation signal and an environmental noise signal; the light source random signal generating device comprises: the light source array is constructed by a plurality of independently luminous light sources, and the driving module is used for driving each independently luminous light source to emit light so as to generate the illumination signal, the electromagnetic radiation signal and the environmental noise signal which are physically and randomly changed; the true random number generation module includes: the system comprises a plurality of sensor set modules for detecting the illumination signal, the electromagnetic radiation signal and the environmental noise signal, a judging module for judging whether at least two groups of effective physical random signals in the three groups of physical random signals are available, and a random number generating module for fusing, scrambling and analyzing the detected physical random signals to generate the true random number.
The invention solves the technical problem by adopting another technical scheme that a method for preparing an eigen state network circuit signal based on a true random number is constructed, and comprises the following steps:
s1, generating a true random number based on the three groups of physical random signals by adopting a random number generator;
s2, generating a control strategy based on the true random number flow and the preset by adopting an SDN controller;
s3, selecting the discrete Hopfield neural network circuit to generate the eigen-state network electric signal based on the control strategy and the true random number flow by adopting a Hopfield neural network device comprising a plurality of discrete Hopfield neural network circuits.
By implementing the intrinsic state network circuit signal preparation system and method based on the true random number, the network circuit signals with high randomness can be rapidly prepared in a large scale. Further, by using three sets of physical random signals and generating random numbers only when at least two sets of physical random signals are normal, a high level of redundancy and entropy for generating each output bit is provided, ensuring that the highest quality random numbers, and thus high quality circuit signals, are produced.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a functional block diagram of a first embodiment of a true random number based eigenstate network circuit signal preparation system of the present invention;
FIG. 2 is a functional block diagram of a random number generator in accordance with a preferred embodiment of the present invention;
FIG. 3 is a schematic diagram of the structure of the light source random signal generating device and the sensor of the random number generator shown in FIG. 2;
FIG. 4 is a functional block diagram of a second embodiment of a true random number based eigenstate network circuit signal preparation system of the present invention;
FIG. 5 is a flow chart of a preferred embodiment of the method for preparing a true random number based eigen state network circuit signal of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
FIG. 1 is a schematic block diagram of a first embodiment of a true random number based eigenstate network circuit signal preparation system of the present invention. As shown in fig. 1, the system for preparing a circuit signal of an eigen state network based on a true random number includes: random number generator 100, Hopfield neural network device 200, internet of things gateway 300, and SDN controller 400. The random number generator 100 is configured to generate a stream of true random numbers based on three sets of physical random signals.
Fig. 2 schematically illustrates a functional block diagram of a preferred random number generator 100 of the present invention. In the preferred embodiment shown in FIG. 2, the random number generator 100 includes a light source random signal generation module 110, a true random number generation module 120, and a verification module 130. As shown in fig. 2, the light source random signal generating module 110 is configured to generate three independent sets of physical random signals. The true random number generation module 120 is configured to generate a stream of true random numbers based on the three sets of physical random signals. The verification module 130 is configured to verify the security and randomness of the true random numbers in the stream of true random numbers using a random statistical verification package.
Preferably, the three independent sets of physical random signals include a light level signal, an electromagnetic radiation signal and an ambient noise signal. In a further preferred embodiment of the present invention, the light source random signal generating module 110 includes: the system comprises a light source array constructed by a plurality of independently luminous light sources, and a driving module used for driving each independently luminous light source to emit light so as to generate the illumination signal, the electromagnetic radiation signal and the environmental noise signal which are physically and randomly changed. The true random number generation module 120 includes: the device comprises a plurality of sensors for detecting the illumination signal, the electromagnetic radiation signal and the environmental noise signal, a judging unit for judging whether at least two groups of physical random signals are effective or not, and a random number generating unit for fusing, scrambling and analyzing the detected physical random signals to generate the true random number stream.
Fig. 3 is a schematic structural diagram of a light source random signal generation module and a sensor of the random number generator shown in fig. 2. As shown in fig. 3, the light source random signal generating module may include 36 LED light sources 61, junction boxes 30, 40, and an embedded industrial personal computer 70, which further includes a signal control card 72 and a signal acquisition card 71. The sensors include a light intensity sensor 51, a sound sensor 53, and a magnetic induction sensor 52. Preferably, 36 LED light sources of five colors of red, green, yellow, white and blue are adopted, and the light source array is configured in a 6-by-6 arrangement.
As shown in fig. 3, the 36 LED light sources 61 can be connected mainly through the junction box 30, and the junction box 30 is connected to the signal control card 72 through the connection cable 10; the specific control action is controlled by the signal control card 72. For example, a PCI slot board PCI-1752U can be used, an isolation digital input channel and an isolation digital output channel can be provided, and the isolation protection voltage can reach 2500 VDC. In addition, all output channels can keep their last output value after the system is restarted, and meanwhile, the PCI-1752U provides a channel freezing function, so that the current output state of each channel can be kept unchanged in operation. The main technical indexes are as follows: 64 isolated digital outputs; output channel high voltage isolation (2500 VDC); 2000VDC ESD protection; a wide input range (5-40 VDC); high sink current on isolated output channels (200 mA maximum/per channel); reading back the output state; maintaining the digital quantity output value when the system is restarted by heat; a channel freeze function. In this embodiment, the 36 LED light sources 61 can be driven in a pseudo-random manner, and the driving function can be customized by the monitoring host, and in combination with the environment, the driving function generates the illumination signal, the electromagnetic radiation signal, and the environmental noise signal that are physically and randomly changed.
As shown in fig. 3, a signal acquisition card 71 may be used to connect the junction box 40 through the connection cable 10, and the junction box 40 is further connected to the illuminance sensor 51, the sound sensor 53 and the magnetic induction sensor 52 through the signal cable 20 to acquire the illuminance signal, the electromagnetic radiation signal and the environmental noise signal. The signal acquisition card 71 can be directly inserted into a PCI slot of an industrial personal computer and is connected with the signal sensors 51-53 through the junction box 40 for signal acquisition.
The signal acquisition card 71 can be selected from the following types: 1. high-precision dynamic signal acquisition card PCIE-1802: the dynamic signal synchronous acquisition card has 8 channels, 24 bits, 216 kS/s/ch. The built-in 4mA/10mA excitation current can be used for measuring Integrated Electronic Piezoelectric (IEPE) sensors, such as sound and vibration signals; 2. multi-channel synchronous sampling multi-function card PCI-1706U: the high-precision universal multifunctional card with 8 channels, 16 bits and 250KS/s is synchronously sampled. It has 8 250KS/s16 bit A/D converters; 3. multichannel scan sampling multifunction card PCI-1716: the multifunctional data acquisition card with 16 channels, 16 bits, 250KS/s and high resolution is provided. It has 1 250KS/s16 bit A/D converter.
The illuminance sensor 51 may be selected from the kunlun coast ZD-6 VBM: the sensor adopts a high-sensitivity photosensitive element as a sensor, and has the characteristics of wide measurement range, good linearity, good waterproof performance, convenience in use and installation, long transmission distance and the like.
The acoustic sensor 53 may be selected from several types: 1. GRAS 40PH/NI 782121-06; the integrated intelligent sensor and the integrated amplifier are powered by IEPE excitation, so that the integrated intelligent sensor and the integrated amplifier are convenient to use. The frequency response range is 10Hz-20kHz, and the SMB interface meets the standard of a class-1 sound level meter; 2. cochingsheng apparatus KSI-308A-213: which is an 1/2 inch electret condenser microphone. The standard preamplifier is supplied with power by a 4mA constant current source (IEPE), the frequency response range is 20Hz-20kHz, and the output is realized by a BNC port. Compared with a preamplifier with polarization voltage, the preamplifier has simple structure and convenient use; 3. a Chengke electronic AWA14423 acoustic sensor + AWA14604 preamplifier; the nickel vibration film and the nickel alloy shell are adopted, special stability treatment is carried out, and the frequency range is wide, the frequency characteristic is good, and the like. The magnetic induction sensor 52 may be, for example, a conway hall type magnetic field sensing module, which uses a linear hall effect sensor for detecting the magnetic induction of the signal source, and has the characteristics of low noise, low power consumption, high precision, inclusion of a thin film resistor, and better temperature stability and accuracy.
In a preferred embodiment of the present invention, when the determining unit determines whether at least two of the three sets of physical random signals are valid, the random number generating unit performs fusion, scrambling and analysis on the detected physical random signals to generate the true random number stream. The verification module 130 may verify the security and randomness of the true random numbers in the stream of true random numbers using a random statistical verification package. In a preferred embodiment of the invention, a comprehensive test is performed using the internationally common standard randomness statistical test kit NIST-STS, so that the generated random numbers can guarantee the highest level of security and randomness.
In this disclosure, the Hopfield neural network device 200 may include a plurality of discrete Hopfield neural network circuits 210, 220, and 230. Although 3 discrete Hopfield neural network circuits are shown in FIG. 1, those skilled in the art will appreciate that the specific number of discrete Hopfield neural network circuits may be constructed according to actual needs.
In the present invention, the SDN controller 400 is configured to generate a control policy based on the true random number stream and a preset. The SDN controller 400 communicates with the random number generator 100 and the Hopfield neural network device 200 through the internet of things gateway 300. The Hopfield neural network device 200 selects the discrete Hopfield neural network circuit 210, 220, 230 to generate the eigen-state network electrical signal based on the control strategy and the stream of true random numbers.
In a preferred embodiment of the present invention, the Hopfield neural network device 200 may include any number of discrete Hopfield neural network circuits. The SDN controller 400 may select an appropriate number of discrete Hopfield neural network circuits according to the number of bits of the eigen-state network electrical signal that needs to be generated. For example, if the Hopfield neural network device 200 includes 50 discrete Hopfield neural network circuits. The number of bits of the eigen state network electric signal to be generated is 32 bits, and 32 suitable discrete Hopfield neural network circuits can be selected according to actual needs to generate the eigen state network electric signal. For example, the priority of the discrete Hopfield neural network circuit may be preset.
In the preferred embodiment, the control strategy includes associating each discrete Hopfield neural network circuit with a true random number, each discrete Hopfield neural network circuit receiving a true random number and outputting an eigen-state network electrical signal. For example, the discrete Hopfield neural network circuit determines whether the received true random number is greater than a set threshold, if so, the discrete Hopfield neural network circuit outputs an eigen state network electrical signal 1, otherwise, the discrete Hopfield neural network circuit outputs an eigen state network electrical signal 0. The set threshold may be a fixed threshold or a dynamic threshold generated based on a random function. For example, the threshold may be selected to be 0.5, and when the true random number received by the discrete Hopfield neural network circuit is greater than 0.5, it outputs the eigen-state network electrical signal 1, otherwise it outputs the eigen-state network electrical signal 0. For another example, the set threshold may be selected to be a dynamic threshold θ generated based on a random function. And if the true random number received by the discrete Hopfield neural network circuit is more than theta, triggering the discrete Hopfield neural network circuit to output an eigen-state network electric signal 1. And if the true random number received by the discrete Hopfield neural network circuit is less than theta, triggering the discrete Hopfield neural network circuit to output an eigen-state network electric signal 0.θ may be generated based on any random function. Any suitable known random function may be employed herein.
The Hopfield neural network device comprises, as the true random number generated by the random number generator is, for example, n bitsn discrete Hopfield neural network circuits, each for generating a circuit signal represented by an n-dimensional vector Y (t) at time t for a one-bit true random number it receives, where Y (t) [ [ Y (t) ] [1(t),Y2(t),...,Yn(t)]TN is an integer greater than zero, Yj(t) (j ═ 1 … … n) takes the value 1 or 0. Thus, the network state has 2nA state; because of Yj(t) (j ═ 1 … … n) can take the value 1 or 0; so that the n-dimensional vector Y (t) has 2nThe seed state is n circuit signal vector states.
In a further preferred embodiment of the invention, said control strategy comprises a one-to-one correspondence of each discrete Hopfield neural network circuit with a true random number, each discrete Hopfield neural network circuit receiving a true random number, and a plurality of discrete Hopfield neural network circuits outputting one said eigenstate network electrical signal. The plurality of discrete Hopfield neural network circuits output an eigen-state network electrical signal 0 or an eigen-state network electrical signal 1 based on the received plurality of true random numbers and a set threshold. For example, 4 true random numbers correspond to the output of an eigen-state network electrical signal, or 2 true random numbers, or 8 true random numbers correspond to the output of an eigen-state network electrical signal. In a preferred embodiment of the present invention, the determination of the output eigenstate network electrical signal 0 or the output eigenstate network electrical signal 1 may be based on a specific value of each true random number and a set threshold. For example, the eigen-state network electrical signal 1 is output only when all the true random numbers are greater than the set threshold, otherwise, the eigen-state network electrical signal 0 is output. Or for example, when half of the true random numbers are larger than the set threshold, the eigen-state network electric signal 1 is output, otherwise, the eigen-state network electric signal 0 is output. For another example, if the sum of all true random numbers is greater than a set threshold, outputting an eigen-state network electrical signal 1, otherwise, outputting an eigen-state network electrical signal 0. Those skilled in the art will appreciate that any suitable criteria may be implemented as desired and as desired by the system.
In a further preferred embodiment of the present invention, each true random number based eigenstate network signal preparation system may comprise an SDN controller, an internet of things gateway and one or more corresponding sets of said Hopfield neural network device 200 and random number generator 100.
The eigen-state network circuit signal preparation system based on the true random number can be used for quickly preparing a large amount of network circuit signals with high randomness. Further, by using three sets of physical random signals and generating random numbers only when at least two sets of physical random signals are normal, a high level of redundancy and entropy for generating each output bit is provided, ensuring that the highest quality random numbers, and thus high quality circuit signals, are produced.
FIG. 4 is a schematic block diagram of a second embodiment of a true random number based eigenstate network circuit signal preparation system of the present invention. Following OSI network protocol stack, adopting software defined network layered control structure, dividing network into physical layer, network layer and control layer. In a preferred embodiment of the present invention, the random number generator 100 and the Hopfield neural network device 200 are disposed in a physical layer. The SDN controller 400 is disposed in the control layer, and the internet of things gateway 300 is disposed in the network layer. The SDN controller 400 communicates with the random number generator 100 and the Hopfield neural network device 200 through the internet of things gateway 300.
In this embodiment, the random number generator 100 may be constructed as described with reference to the embodiments shown in FIGS. 1-3. This will not be described in detail.
In the preferred embodiment of the present invention, the Hopfield neural network device 200 utilizes the characteristic that true random numbers correspond to signal generating bits one to one, and adopts a plurality of discrete Hopfield neural network circuits to realize the parallelism of multiple signal generating bits. For each bit of true random number, a discrete Hopfield neural network circuit is adopted to correspond to each bit of signal generation bit, and by utilizing the binary property of the discrete Hopfield neural network, corresponding output discrete values 1 and 0 respectively represent that the neuron is in an activation state and a suppression state, namely 1 or 0 steady state of the signal generation bit.
In the embodiment, each discrete Hopfield neural network circuit is corresponding to a true random number one by one, and each discrete Hopfield neural network circuit receives a true random number and outputs an eigen-state network electric signal. In this embodiment, the discrete Hopfield neural network circuit determines whether the received true random number is greater than a set threshold, and if so, the discrete Hopfield neural network circuit outputs an eigen-state network electrical signal 1, otherwise, the discrete Hopfield neural network circuit outputs an eigen-state network electrical signal 0. Here, a dynamic threshold, i.e., a threshold θ based on a random function, is used, so that a true random number smaller than θ in the true random number stream outputs an eigen state network electrical signal 0 through the discrete Hopfield neural network circuit, and a true random number smaller than θ outputs an eigen state network electrical signal 1 through the discrete Hopfield neural network circuit.
For the generation of a parallel vector of n signal generation bits, a Hopfield neural network device comprising n discrete Hopfield neural network circuits may be used. Each discrete Hopfield neural network circuit is configured to generate an n-dimensional vector Y (t) representing an eigenstate network electrical signal at time t for a one-bit true random number received by the circuit, wherein Y (t) ═ Y1(t),Y2(t),...,Yn(t)]TN is an integer greater than zero, Yj(t) (j ═ 1 … … n) takes the value 1 or 0. Thus, the network state has 2nA state; because of Yj(t) (j ═ 1 … … n) can take the value 1 or 0; so that the n-dimensional vector Y (t) has 2nAnd the n states are positive and negative vector states of the network electric signal of the eigenstates.
Further, the network layer is mainly responsible for network communication. Besides network connection and management, the functions of the system also comprise field processing, and the system can ensure the local survival of network circuit signal preparation operation. And collecting state maintenance data of each device of the physical layer. In addition, protocol translation is also an important function of this layer. Protocol conversion is required to be carried out on the layer of gateway, and data is uniformly loaded on the IP network and transmitted outwards. The internet of things gateway 300 includes an exchange routing and ICT convergence gateway, so the internet of things gateway 300 disposed in the network layer may include an exchange routing and ICT convergence gateway, which supports a plurality of types of interfaces, for example, physical interfaces supporting a plurality of industrial standards, and a plurality of types of supported protocols, and has local intelligence, that is, the gateway is required to have local computing, edge decision, and local survival capabilities, so as to ensure that the connected eigen-state-based circuit signal preparation apparatus may encode the instruction requirements of the internet of things gateway 300 according to the SDN controller 400, thereby preparing a specific initial state of the N-information circuit signal. The internet of things gateway 300 can enable different types of network circuit signal preparation applications such as local computation, nearby observation, edge survival maintenance and the like to be operated on the gateway through the SDN/NFV technology, so that low-delay service and local survival are possible.
Preferably, the control layer may further include a network management function, a computing resource management function, an application management and data subscription and release function, and a physical circuit control protocol loading function, wherein the network management function includes unified management and authentication of a physical layer terminal and a gateway. The computing resource management function may enable management and allocation of computing resources for the internet of things gateway 300. The application management may implement downloading and updating of gateway applications such as light, magnetic field, working strategies for noise signal generation, device state maintenance mechanisms, etc. in the random number generator 100. The data subscribing and issuing function comprises the steps of processing state maintenance data of the device and observing signal simulation operation processes of the information circuit. The physical circuit control protocol loading function comprises the steps of providing abundant illumination, magnetic fields and noise signal control protocols to meet the docking requirements of a physical layer, and dynamically loading the protocols according to needs.
Under the condition that the demand of large-scale data analysis on super-strong computing power is continuously increased, the invention provides a multi-bit eigenstate network circuit signal parallelization preparation system which can adopt a virtualization technology to perform segmentation organization on computing service resources of a single circuit signal simulation preparation node and prepare and form a superposition image state facing communication, storage and service environments based on a computing system of SDN/NFV. The eigen-state network circuit signal preparation system based on the true random number can adopt a distributed parallel computation mode to synchronously copy and migrate the service computation flow to each image state to participate in independent computation, thereby realizing high-speed parallel computation support. The eigen-state network circuit signal preparation system based on the true random number ensures that the process flow of the eigen-state network circuit signal preparation can be realized in a software definition mode. Based on a software definition mechanism, a specified reconstruction strategy set can be adopted to superpose the independent calculation result of each image state, and the result output of a network system is completed cooperatively.
FIG. 5 is a flow chart of a preferred embodiment of the method for preparing a true random number based eigen state network circuit signal of the present invention. As shown in fig. 5, in step S1, a true random number is generated based on the three sets of physical random signals using a random number generator. In step S2, a control strategy is generated based on the true random number stream and a preset using an SDN controller. In step S3, a Hopfield neural network device comprising a plurality of discrete Hopfield neural network circuits is employed to select the discrete Hopfield neural network circuits to generate the eigenstate network electrical signal based on the control strategy and the stream of true random numbers.
In a further preferred embodiment of the present invention, the random number generator 100 and the Hopfield neural network module 200 may be constructed with reference to the embodiments shown in FIGS. 1-4. A further true random number based eigen-state network circuit signal preparation method of the present invention may also be constructed with reference to the embodiments shown in fig. 1-4.
The preparation of the eigen-state network circuit signal of 32 bits is explained as follows. The signal is represented in binary, 4 segments as follows:
11110000、11101110、11001100、10001000
because the support of fast parallel computing is considered, the generation of the multi-bit intrinsic signal is different from the generation of the same-bit signal of the traditional von Neumann computer, and the signals generated by the system are simultaneously generated in parallel, namely 32-bit signals are generated in parallel at the same time, and the 32-bit signals are not generated in a serial instruction mode of the CPU of the existing system.
The physical layer of the system receives the control instruction and controls the physical circuit to start generating the true random number by adopting a rule of generating a circuit signal every 6 seconds. The relay module or the control board independently drives a 6 x 6 light source array consisting of 36 LEDs of five colors of red, green, yellow, white and blue, and the sensor receives three independent groups of physical random signals generated by continuous light sources: a light level signal, an electromagnetic radiation signal and an ambient noise signal. And judging whether at least two groups of the three groups of physical random signals are effective, and if the signals are effective, generating a random number sequence according to the detected physical random signals. If the signal is invalid, the generation of the true random number sequence is stopped. And finally, verifying the safety and the randomness of the true random numbers in the true random number sequence by adopting a random statistical test packet specified by NIST SP800-22 standard, and finishing the generation of the true random number sequence. The random number generator is connected with a hopfield neural network.
And the SDN judges according to the generated true random number sequence, different random numbers correspond to different Hopfield neural networks according to a control strategy, and simultaneously, the SDN controls the network input and output of corresponding bits of different bits. In this embodiment, the SDN controller determines an output of each bit corresponding to each bit according to an eigen state of a 32-bit signal, generates a true random number positive and negative direction generation control measure as a control command to control the physical layer, and the control command issues a control strategy for forming an eigen state network circuit signal to the physical layer through the network layer. For example, if the value of the random number in the random sequence is larger than theta, the corresponding Hopfield neural network circuit in the Hopfield neural network device is triggered to output a1 signal, and if the value of the random number in the random sequence is smaller than theta, the corresponding Hopfield neural network circuit in the Hopfield neural network device outputs the signal. Of course, it is also possible to select 4 Hopfield neural network circuits as inputs, but to output only one eigen-state network circuit signal, as described above.
In a preferred embodiment of the present invention, a plurality of the above-mentioned true random number-based eigenstate network circuit signal preparation systems can be included in the network. And after the preparation of the eigen-state network circuit signals is finished, the interconnection of all eigen-state network circuit signal preparation systems based on true random numbers is realized through a network layer. All eigen-state network circuit signals are passed up to the control layer. A physical gateway and an SDN controller may be used to connect multiple sets of random number generators and corresponding Hopfield neural network devices, so as to ensure cooperation between each random number generator and the corresponding Hopfield neural network device, and implement parallel generation of signals.
The SDN controller uniformly processes the intrinsic state network circuit signals transmitted by the network layer for physical layer maintenance and signal position observation. And management control is carried out on the network, the control component, the measurement component and the computing resource of the lower layer. Through this step, the SDN controller may measure and check parallel bits or vectors of the output signal, and ensure the correctness of the output eigen-state signal.
And finally, synthesizing the 32-bit random number and the final output of the corresponding Hopfield neural network through the global management control of the SDN to obtain the final stable eigenstate result of the 32-bit signal, and flexibly realizing the initial state of the software-defined multi-bit eigenstate network circuit signal. The SDN controller controls each Hopfield circuit to generate a circuit signal every 6 seconds through a circuit signal generation strategy sent by a receiving and connecting gateway, and the system can generate 320 eigenstate network circuit signals within one minute.
The eigen-state network circuit signal preparation system based on the true random number can be used for quickly preparing a large amount of network circuit signals with high randomness. Further, by using three sets of physical random signals and generating random numbers only when at least two sets of physical random signals are normal, a high level of redundancy and entropy for generating each output bit is provided, ensuring that the highest quality random numbers, and thus high quality circuit signals, are produced.
Accordingly, the present invention can be realized in hardware, software, or a combination of hardware and software. The present invention can be realized in a centralized fashion in at least one computer system, or in a distributed fashion where different elements are spread across several interconnected computer systems. Any kind of computer system or other apparatus adapted for carrying out the methods of the present invention is suited. A typical combination of hardware and software could be a general purpose computer system with a computer program that, when being loaded and executed, controls the computer system such that it carries out the methods described herein.
The present invention may also be implemented by a computer program product, comprising all the features enabling the implementation of the methods of the invention, when loaded in a computer system. The computer program in this document refers to: any expression, in any programming language, code or notation, of a set of instructions intended to cause a system having an information processing capability to perform a particular function either directly or after either or both of the following: a) conversion to other languages, codes or symbols; b) reproduced in a different format.
While the invention has been described with reference to specific embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted without departing from the scope of the invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from its scope. Therefore, it is intended that the invention not be limited to the particular embodiment disclosed, but that the invention will include all embodiments falling within the scope of the appended claims.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (6)
1. A preparation system of eigen state network circuit signals based on true random numbers is characterized by comprising the following steps:
a random number generator for generating a stream of true random numbers based on the three sets of physical random signals;
an SDN controller for generating a control strategy based on the true random number flow and a preset;
a Hopfield neural network device comprising a plurality of discrete Hopfield neural network circuits;
an instrumented gateway through which the SDN controller communicates with the random number generator and the Hopfield neural network device;
wherein the Hopfield neural network device selects the discrete Hopfield neural network circuit to generate an eigen-state network electrical signal based on the control strategy and the true random number stream; the control strategy comprises the steps that each discrete Hopfield neural network circuit corresponds to a true random number one by one, and each discrete Hopfield neural network circuit receives the true random number and outputs an eigen-state network electric signal;
the Hopfield neural network device comprises n discrete Hopfield neural network circuits, each for generating a circuit signal represented by an n-dimensional vector Y (t) at time t for a one-bit true random number it receives, wherein Y (t) ═ Y (t)1(t),Y2(t),...,Yn(t)]TN is an integer greater than zero, Yj(t) takes the value 1 or 0, where j =1 … … n.
2. A preparation system of eigen state network circuit signals based on true random numbers is characterized by comprising the following steps:
a random number generator for generating a stream of true random numbers based on the three sets of physical random signals;
an SDN controller for generating a control strategy based on the true random number flow and a preset;
a Hopfield neural network device comprising a plurality of discrete Hopfield neural network circuits;
an instrumented gateway through which the SDN controller communicates with the random number generator and the Hopfield neural network device;
wherein the Hopfield neural network device selects the discrete Hopfield neural network circuit to generate an eigen-state network electrical signal based on the control strategy and the true random number stream;
the discrete Hopfield neural network circuit judges whether the received true random number is larger than a set threshold value, if so, the discrete Hopfield neural network circuit outputs an eigen state network electric signal 1, otherwise, the discrete Hopfield neural network circuit outputs an eigen state network electric signal 0; the set threshold is a fixed threshold or a dynamic threshold generated based on a random function.
3. A preparation system of eigen state network circuit signals based on true random numbers is characterized by comprising the following steps:
a random number generator for generating a stream of true random numbers based on the three sets of physical random signals;
an SDN controller for generating a control strategy based on the true random number flow and a preset;
a Hopfield neural network device comprising a plurality of discrete Hopfield neural network circuits;
an instrumented gateway through which the SDN controller communicates with the random number generator and the Hopfield neural network device;
wherein the Hopfield neural network device selects the discrete Hopfield neural network circuit to generate an eigen-state network electrical signal based on the control strategy and the true random number stream;
the control strategy comprises that each discrete Hopfield neural network circuit corresponds to a true random number one by one, each discrete Hopfield neural network circuit receives a true random number, and a plurality of discrete Hopfield neural network circuits output an eigen-state network electric signal; the plurality of discrete Hopfield neural network circuits output an eigen-state network electrical signal 0 or an eigen-state network electrical signal 1 based on the received plurality of true random numbers and a set threshold.
4. The true random number based eigenstate network circuit signal preparation system of any of claims 1-3, wherein the random number generator further comprises:
the light source random signal generating module is used for generating three groups of independent physical random signals;
a true random number generation module for generating a true random number stream based on the three sets of physical random signals;
and the verification module is used for verifying the safety and the randomness of the true random number by adopting a random statistical verification package.
5. The true random number based eigenstate network circuit signal preparation system of claim 4, wherein the three independent sets of physical random signals include a light level signal, an electromagnetic radiation signal, and an ambient noise signal; the light source random signal generating device comprises: the light source array is constructed by a plurality of independently luminous light sources, and the driving module is used for driving each independently luminous light source to emit light so as to generate the illumination signal, the electromagnetic radiation signal and the environmental noise signal which are physically and randomly changed; the true random number generation module includes: the system comprises a plurality of sensor set modules for detecting the illumination signal, the electromagnetic radiation signal and the environmental noise signal, a judging module for judging whether at least two groups of effective physical random signals in the three groups of physical random signals are available, and a random number generating module for fusing, scrambling and analyzing the detected physical random signals to generate the true random number.
6. A method for preparing an eigen state network circuit signal based on a true random number is characterized by comprising the following steps:
s1, generating a true random number based on the three groups of physical random signals by adopting a random number generator;
s2, generating a control strategy based on the true random number flow and the preset by adopting an SDN controller;
s3, selecting the discrete Hopfield neural network circuit to generate an eigen-state network electric signal by adopting a Hopfield neural network device comprising a plurality of discrete Hopfield neural network circuits based on the control strategy and the true random number flow;
the control strategy comprises the steps that each discrete Hopfield neural network circuit corresponds to a true random number one by one, and each discrete Hopfield neural network circuit receives the true random number and outputs an eigen-state network electric signal; the Hopfield neural network device comprises n discrete Hopfield neural network circuits, each for generating a circuit signal represented by an n-dimensional vector Y (t) at time t for a one-bit true random number it receives, wherein Y (t) ═ Y (t)1(t),Y2(t),...,Yn(t)]TN is an integer greater than zero, Yj(t) has a value of 1 or 0, wherein j =1 … … n; or
The control strategy comprises that each discrete Hopfield neural network circuit corresponds to a true random number one by one, each discrete Hopfield neural network circuit receives a true random number, and a plurality of discrete Hopfield neural network circuits output an eigen-state network electric signal; the plurality of discrete Hopfield neural network circuits output eigen-state network electric signals 0 or output eigen-state network electric signals 1 based on the plurality of received true random numbers and a set threshold; or
The control strategy comprises that each discrete Hopfield neural network circuit corresponds to a true random number one by one, each discrete Hopfield neural network circuit receives a true random number, and a plurality of discrete Hopfield neural network circuits output an eigen-state network electric signal; the plurality of discrete Hopfield neural network circuits output an eigen-state network electrical signal 0 or an eigen-state network electrical signal 1 based on the received plurality of true random numbers and a set threshold.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010618163.1A CN111860774B (en) | 2020-06-30 | 2020-06-30 | True random number-based eigen state network circuit signal preparation system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010618163.1A CN111860774B (en) | 2020-06-30 | 2020-06-30 | True random number-based eigen state network circuit signal preparation system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111860774A CN111860774A (en) | 2020-10-30 |
| CN111860774B true CN111860774B (en) | 2021-10-22 |
Family
ID=72989869
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010618163.1A Active CN111860774B (en) | 2020-06-30 | 2020-06-30 | True random number-based eigen state network circuit signal preparation system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111860774B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115811397B (en) * | 2022-11-21 | 2023-08-04 | 北京神州安付科技股份有限公司 | High-safety server cipher machine |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019055847A1 (en) * | 2017-09-15 | 2019-03-21 | President And Fellows Of Harvard College | Quantum artificial neural networks |
| CN110333842A (en) * | 2019-06-28 | 2019-10-15 | 深圳市永达电子信息股份有限公司 | A kind of true random number generation method and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108664234A (en) * | 2018-01-17 | 2018-10-16 | 北京智芯微电子科技有限公司 | Real random number generator |
-
2020
- 2020-06-30 CN CN202010618163.1A patent/CN111860774B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019055847A1 (en) * | 2017-09-15 | 2019-03-21 | President And Fellows Of Harvard College | Quantum artificial neural networks |
| CN110333842A (en) * | 2019-06-28 | 2019-10-15 | 深圳市永达电子信息股份有限公司 | A kind of true random number generation method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111860774A (en) | 2020-10-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Neural network renormalization group | |
| Zhou et al. | Tgl: A general framework for temporal gnn training on billion-scale graphs | |
| CN110333842B (en) | True random number generation method and system | |
| Killoran et al. | Converting nonclassicality into entanglement | |
| CN111860774B (en) | True random number-based eigen state network circuit signal preparation system and method | |
| Sahani et al. | Deep convolutional stack autoencoder of process adaptive VMD data with robust multikernel RVFLN for power quality events recognition | |
| CN109239576B (en) | High-speed optical communication chip test system and method | |
| CN111832732B (en) | Digital quantum bit preparation device and method based on eigenstates | |
| Abshire et al. | Capacity and energy cost of information in biological and silicon photoreceptors | |
| JP2001523829A (en) | Method and apparatus for detecting and matching signal waveforms that can be changed from part to part within a monotonous and constant trajectory | |
| CN111884792A (en) | System and method for preparing eigen-state network circuit signal based on memristor chaotic sequence | |
| JP2017108220A (en) | Data collection device and optical transmission system | |
| Dheeraj et al. | Detecting Deepfakes Using Deep Learning | |
| US11894857B2 (en) | Readout circuit, signal quantizing method and device, and computer device | |
| Zheng et al. | Nonadaptive group testing based fiber sensor deployment for multiperson tracking | |
| WO2018142694A1 (en) | Feature amount generation device, feature amount generation method, and program | |
| Alsulami et al. | Improvement on a traffic data generator for networking AI algorithm development | |
| US11507803B2 (en) | System for generating synthetic digital data for data multiplication | |
| CN114494484A (en) | Training method of data recognition model, data recognition method, device and equipment | |
| Farnbauer-Schmidt et al. | Combining the concepts of semantic data integration and edge computing | |
| JP3717478B2 (en) | Bit error test method and apparatus | |
| CN202102484U (en) | Coincidence counting system for detection experiment of entangled photons | |
| Gaisbauer et al. | Unified communication framework | |
| Tambova et al. | Colored Petri Nets Model for Network Based FTI Systems. | |
| Presti et al. | A grid enabled parallel hybrid genetic algorithm for SPN |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |