Device and method for generating random bit sequence

Abstract

An apparatus (200) for generating a random bit sequence comprises: at least one silicon photomultiplier sensor (201) configured to generate a sequence (S1) of endogenous random current pulses due to self-amplification of thermally generated charge carriers driven by impact ionization, the at least one silicon photomultiplier sensor (201) being subject to thermally generated charge carriers; a data processing unit (202) configured to receive the sequence (S1) of endogenous random current pulses and determine a random bit sequence (S2) to be provided to an end user (EU) based on the sequence (S1) of endogenous random current pulses received from the at least one silicon photomultiplier sensor (201).

Description

Technical Field

[0001] The present invention relates to random number generation, i.e., devices and methods for generating a sequence of random bits (bit: bit). Background Art

[0002] It is known that random number generation can be based on algorithms or on the observability related to unpredictable natural phenomena. The former is implemented in software or firmware, and the latter requires a hardware system for information collection and a method for processing it to extract a series of random numbers.

[0003] Algorithm-based random generation certainly benefits from computing power and optimal programming to achieve extremely high data rates, i.e., far exceeding 10 Gb / s.

[0004] However, it has fundamental and irreducible drawbacks:

[0005] - The algorithm is deterministic. Therefore, the generated sequence is pseudo-random. The sequence may have a period that meets most requirements, but its random characteristics will be irreducibly limited;

[0006] - The generated sequence depends on a digital seed to initialize the program. Therefore, accessing the sequence means accessing the entire series of generated numbers.

[0007] Hardware-based random number generation is based on natural phenomena, either described by classical physics or based on the quantum properties of nature.

[0008] The classical description is deterministic.

[0009] Even if it is possible to assume the complexity of the system or its chaotic nature to provide a basis for the actual unpredictability that occurs, the essence of natural phenomena is that once the initial conditions are known or reproduced in a controlled manner, the dynamics of the system are well-defined.

[0010] On the other hand, the phenomena at the quantum level are inherently random and thus unpredictable. For this reason, they are an ideal basis for true random number generation (hereinafter referred to as "TRNG").

[0011] Random number generators are applied in many fields, such as: computer security and cryptography; Internet of Things (IoT) devices, where some of the services provided by the devices rely on confidentiality to protect the privacy of end-users and avoid intrusion and hacking; digital simulation of complex phenomena in science, industry (aerodynamics, thermodynamics, and manufacturing), economics, and sociology, where it has been shown that the quality of random numbers is crucial for the reliability of simulation results; optimization in communication across congested networks; development of online platforms.

[0012] Historically, the first quantum random number generators were based on unstable radionuclides, whose decay emits alpha, beta, or gamma particles.

[0013] The emissions occur in an unpredictable manner, and the number of decays within a pre-defined time window follows a Poisson distribution.

[0014] In other words, the time elapsed between two consecutive events follows a probability density function of exponential decay, whose decay constant depends on the isotope used and its radioactivity. The pulses are statistically independent and uncorrelated, and random bit generation can be obtained in different ways.

[0015] Radioactive decay remains a very robust and rather simple way to obtain a random bit stream today.

[0016] However, they have obvious health protection and safety issues, which prevent their large-scale adoption.

[0017] In addition, the characteristics of the particle detectors, especially their dead time and radiation damage, limit the available flux and undermine the stability.

[0018] Finally, even in a dedicated location, the handling and storage of the radiation source make the system economically uncompetitive.

[0019] So far, most quantum random number generators rely on weak light sources and detectors with single-photon sensitivity, using various setups and arrangements.

[0020] Figure 1 An example illustration of a quantum random number generator is schematically reported and denoted by reference numeral 100.

[0021] Generator 100 includes a pulsed light source 101, a beam splitter 102, a first sensor D0, and a second sensor D1.

[0022] The pulsed light source 101 emits single photons that pass through the beam splitter 102.

[0023] Assuming that the pulsed light source 101 emits single photons and the beam splitter 102 is perfect, for each pulse, the first sensor D0 and the second sensor D1 have an equal chance of receiving the photon.

[0024] As long as their efficiencies are the same, a bit with a value of 0 is generated when a photon is detected by the first sensor D0, and a bit with a value of 1 is generated when a photon is detected by the second sensor D1.

[0025] Figure 1 The setup of the quantum random generator has a series of weaknesses that may affect the quality of the extracted bit sequence.

[0026] In fact, a standard light source emits a random number of photons per pulse, typically following a Poisson distribution. Whether the randomness can be a critical value or not, in the Figure 1 setting, it represents a limiting factor. The requirement to reduce the probability of having more than one emitted photon to the level of 10 -8 demands that the average number of emitted photons per pulse does not exceed a few 10 -4 , affecting the achievable rate.

[0027] In addition, any imperfection in the beam splitter 102 affects the probabilities of generating 0 and 1. The same occurs unless the first sensor D0 and the second sensor D1 are identical in terms of photon detection probability and stability against temperature or operating voltage variations.

[0028] It should be noted that the weaknesses in the instrumentation and the setting can be overcome by properly engineering the system and implementing a method for balancing the probabilities of generating an unpredictable series of bit values through a post - processing algorithm.

[0029] Even so, high - quality uniform random bits in megabits per second (Mbps) can be achieved, but at the cost of increased system complexity and lower efficiency in terms of useful bits per random event.

[0030] According to other solutions belonging to the prior art, examples of the setting rely on single - photon - sensitive detectors, possibly photon - number resolving, to detect, time - stamp, and record uncorrelated single photons emitted by a attenuated laser source or a light - emitting diode (LED).

[0031] In particular, a quantum random number generator that utilizes the arrival time of photons is based on time - stamping the detection of single photons from an LED source, detected with a photomultiplier tube (PMT), mimicking the procedure implemented for random pulses by a radioactive source in a Geiger - Müller or silicon - based detector.

[0032] According to another solution of the prior art, the photomultiplier tube has been replaced by a single - photon avalanche photodiode (SPAD), and the overall setting has been integrated on a single chip.

[0033] A variant of the same principle is based on counting the pulses originating from a light source within a predefined time interval and relying on the Poisson characteristics of the underlying distribution.

[0034] According to solutions belonging to the prior art, random events correspond to the detection or absence of light by a pulsed light source in a single - photon state. Randomness extraction is based on aggregating random events in blocks and applying an algorithm that requires resources to increase polynomially with the block length, implemented in a field - programmable gate array (FPGA).

[0035] According to another solution of the prior art, an array of single-photon avalanche photodiodes (SPADs) is irradiated by an LED source that emits bursts of photons at fully tuned time intervals, giving each pixel a 50% chance of detecting a photon.

[0036] However, regardless of whether the design of this setup is definitely conducive to its parallel implementation, in principle, it allows random bit generation at Gbps rates, but its main weakness lies in the requirement for "perfect" time tuning of the pulse duration.

[0037] Another solution of the prior art, instead, is based on generating two random bit strings according to the statistical properties of the Poisson distribution in the number of photons detected per pulse by a silicon photomultiplier - i.e., an array of single-photon avalanche photodiodes (SPADs) with a conventional output.

[0038] However, again, the principle is significant, but the implementation is rather weak, requiring perfect separation in the distribution of detected photons / pulses, an extremely controlled system, and precise calibration before and during sequence generation.

[0039] In summary, all solutions based on mimicking particle detection by using a light pulse radiation source present:

[0040] - Complexity in setup due to the characteristics of the light source and the requirements of the dual-source detector system;

[0041] - Lack of robustness associated with the requirement for extreme stability against temperature and voltage variations;

[0042] - In some cases, a low random bit rate extracted per event. Summary of the Invention

[0043] The object of the present invention is to conceive and provide a device for generating a random bit sequence that allows at least partially solving the defects referred to above with respect to the prior art, having high reliability and low setup complexity, ensuring high stability against temperature and voltage variations, and a relatively high random bit rate extracted per event.

[0044] Such an object is achieved by a device for generating a random bit sequence according to an embodiment of the present invention.

[0045] The preferred embodiments of the device are defined in a device for generating a random bit sequence according to a further embodiment of the present invention.

[0046] In addition, the object of the present invention is a method for generating a random bit sequence. Brief Description of the Drawings

[0047] In the following description, other features and advantages of the device and method according to the present invention will become apparent. The following description shows, by way of indicative and non-limiting examples, preferred embodiments with reference to the accompanying drawings, in which:

[0048] Figure 1 A random number generator belonging to the prior art is schematically shown by a block diagram;

[0049] - Figure 2 The trend of the pulse sequence varying with time is schematically shown by a timing diagram, and the pulse sequence can be generated by a device for producing a random bit sequence according to the present invention;

[0050] - Figure 3 A device for generating a random bit sequence according to an embodiment of the present invention is schematically shown by a block diagram;

[0051] - Figure 4 A device for generating a random bit sequence according to an embodiment of the present invention is schematically shown by a block diagram, and

[0052] - Figure 5 A method for generating a random bit sequence according to an embodiment of the present invention is schematically shown by a block diagram. Detailed Embodiments

[0053] With reference to the foregoing drawings, according to the present invention, a device 200 for generating a random bit sequence will now be described, which will also be briefly described hereinafter.

[0054] Specifically with reference to Figure 3 and Figure 4 the embodiments, the device 200 includes at least one silicon photomultiplier sensor 201, which is configured to generate a sequence S1 of endogenous random current pulses due to the self-amplification of thermally generated charge carriers driven by impact ionization, and the at least one silicon photomultiplier sensor 201 can be subjected to the thermally generated charge carriers.

[0055] In this regard, it should be observed that the applicant has noted the possibility of generating random bits by time-stamping the pulses self-amplified due to randomly generated charge carriers in an array of p-n junctions (cells) operating above the breakdown voltage, i.e., a device known as a silicon photomultiplier or multi-pixel photon counter.

[0056] This endogenous mechanism is expected to overcome the complexity and weaknesses of existing random number generators based on quantum effects that rely on exogenous pulse sources, either radioactive sources or photon sources.

[0057] The quantum properties of energy bands in semiconductor devices, the distribution of electrons among energy levels according to Fermi-Dirac statistics, and the strong electric field effects in microelectronic devices provide this mechanism.

[0058] In fact, in silicon and other indirect semiconductor materials, trap-assisted thermally driven random generation and recombination of free carriers are dominant, and it is the physical phenomenon underlying the generation of current in the depletion region of a p-n junction.

[0059] Furthermore, if the junction operates in the avalanche regime, this mechanism is the cause of the occurrence of random pulses.

[0060] Regardless of the mechanism, the applicant notes that the key point lies in the high density of potential carriers. The random occurrence of bringing potential carriers to the conduction band, together with the random probability of causing avalanche breakdown, results in a series of independent pulses. These independent pulses are expected to follow a Poisson distribution. Therefore, the pulses are uncorrelated and their appearance is unpredictable.

[0061] Currently, various manufacturers offer silicon photomultipliers (SiPMs) on the market, with a density of up to 40,000 cells / mm2

[40] , an area exceeding 6×6 mm2, and the existing interconnect technologies allow the construction of mosaics of evenly balanced sensors over large areas.

[0062] In an SiPM, the strong electric field in the junction volume causes impact ionization to multiply charges, where, at a bias voltage not exceeding 70 V, the gain can be up to 10 6 。

[0063] In particular, an SiPM operates essentially in a Geiger-Müller regime with limited quenching, such that a single charge carrier can trigger a pulse with a probability approaching 100% at a few volts above the breakdown voltage.

[0064] SiPMs are commercialized as single-photon sensitive and photon-number resolving detectors, and their extremely high sensitivity is due to the avalanche triggered by a single charge carrier released by a photon.

[0065] However, the avalanche development is blind to the mechanism for generating primary carriers: it can be a photon, or any endogenous random process related to temperature or electric field mapping.

[0066] This is the principle underlying the present invention, in that an SiPM is encapsulated in complete darkness, random-initiated pulses are identified, time-tagged, and the sequence of pulses is transformed into a series of bits. The amplitude of the pulses - millions of electrons in a few tens of nanoseconds - makes their identification robust and error-free; the time development of the avalanche and the leading edge of the nanosecond-scale signal make the time tagging extremely precise; the endogenous generation mechanism makes the process robust and resistant to temperature variations at a predicted rate of change without destroying the randomness.

[0067] The pulse rate reaches 1 MHz / mm2 at room temperature a few volts above the breakdown voltage, offering the possibility of designing a very compact device for engineering.

[0068] Figure 2 An exemplary illustration of randomly generated pulses E1, E2, E3, E4 in at least one silicon photomultiplier sensor 201 of an embodiment of a SiPM such as Figure 3 and Figure 4 is shown.

[0069] After pulse identification, the arrival time t or is recorded relative to the reference time origin t i , i = 1, 2, 3, 4... N, as shown in the lower panel of Figure 2 .

[0070] If the pulses are uncorrelated and independent, then in the intervals (dT i , dT j ), where dT i = t i+1 - t i , dT i is as likely to be longer or shorter than dT j , i.e.:

[0071] P(dT i > dTj) = P(dT i < dT j ) = 0.5.

[0072] Based on this general property of the memoryless event sequence, bits are extracted according to the following procedure:

[0073] If dT i > dT i+2 , then the i-th bit in the sequence is set to 1;

[0074] If dT i < dT i+2 , then the i-th bit in the sequence is set to 0;

[0075] If dT i = dT i+2 , then no bit is generated.

[0076] In the exemplary illustration of Figure 2 :

[0077] dT 1 = t 2 – t 1 ;

[0078] dT 2 = t3 –t 2 ;

[0079] dT 3 = t 4 –t 3 ;

[0080] dT 1 <dT 3 If so, the first bit in the sequence is set to 0.

[0081] This program ensures a bit extraction efficiency of up to 50%, which may be reduced due to the number of equal time intervals.

[0082] In addition, it is unbiased (including the potential impact of timestamping by a continuous clock), and does not require a processing algorithm that would reduce the effective bit rate.

[0083] At least one silicon in the silicon photomultiplier sensor 201 is configured to collect entropy, and as already described above, generate a sequence of endogenous random current pulses through self-amplification (e.g., gain 106), i.e., avalanche, of thermally generated charge carriers driven by impact ionization.

[0084] Go to Figure 3 and Figure 4 In the implementation of, the device 200 further includes a data processing unit 202, which is configured to receive the sequence S1 of the endogenous random current pulses and determine a random bit sequence S2 to be provided to the end user EU based on the sequence S1 of the endogenous random current pulses received from the at least one silicon photomultiplier sensor 201.

[0085] Examples of the end user EU are a key generation device, a network random coding system.

[0086] According to Figure 3 In the implementation shown in, the data processing unit 202 of the device 200 includes a signal conditioning module 203 configured to receive the sequence S1 of the endogenous random current pulses.

[0087] The signal conditioning module 203 of the device 200 is configured to determine a logic stop signal STP for each current pulse in the sequence S1 of the endogenous random current pulses received from the at least one silicon photomultiplier sensor 201.

[0088] The signal conditioning module 203 can be regarded as an analog front end (AFE) module.

[0089] According to Figure 3In an embodiment, the data processing unit 202 of the device 200 further includes a time-to-digital converter module 204, which is configured to receive a logic start signal STR and each logic stop signal STP determined by the signal conditioning module 203.

[0090] The time-to-digital converter module 204 is further configured to measure the time interval between the arrival time of each logic stop signal STP and the arrival time of the logic start signal STR.

[0091] According to Figure 3 In an embodiment, the data processing unit 202 of the device 200 further includes a control module 205, which is configured to provide the logic start signal STR to the time-to-digital converter module 204 when the device 200 is powered on.

[0092] According to this embodiment, the control module 205 is further configured to process each time interval measured by the time-to-digital converter module 204 and generate a random bit sequence S2 to be provided to the end user EU based on each processed time interval.

[0093] According to an embodiment, the control module 205 is a system-on-chip (SoC) module.

[0094] The control module 205 is configured to communicate with other modules of the device 200 through a serial peripheral interface (SPI) and / or an inter-integrated circuit (I 2 C line).

[0095] According to Figure 3 In the embodiment shown, the device 200 further includes a high-voltage power supply module 206, which is configured to provide a high-voltage power supply to at least one silicon photomultiplier sensor 201.

[0096] The high-voltage power supply module 206 is further configured to control the dependence of the operating voltage and the excess bias voltage of at least one silicon photomultiplier sensor 201 relative to the breakdown voltage value.

[0097] According to an embodiment, in combination with the previous embodiment, the control module 205 is configured to control at least one silicon photomultiplier sensor 201 via the high-voltage power supply module 206.

[0098] According to this embodiment, the control module 205 is further configured to further control the signal conditioning module 203 and the time-to-digital converter module 204.

[0099] As described above, the signal conditioning module 203 can be regarded as an analog front-end (AFE) module.

[0100] In this regard, according to Figure 4In the illustrated embodiment, the signal conditioning module 203 includes a current-voltage conversion module 401 configured to convert each current pulse in the sequence S1 of endogenous random current pulses into a corresponding voltage pulse.

[0101] An example of the current-voltage conversion module 401 is a transimpedance amplifier.

[0102] According to Figure 4 In the illustrated embodiment, the signal conditioning module 203 further includes a level adapter module 402 configured to condition the level of each voltage pulse.

[0103] The level adapter module 402 is downstream of the current-voltage conversion module 401.

[0104] An example of the level adapter module 402 is a module for canceling signals.

[0105] According to this embodiment, the signal conditioning module 203 further includes a comparator module 403 configured to compare each voltage pulse with a tunable threshold.

[0106] The comparator module 403 is further configured to provide a logic voltage level output whenever the threshold is exceeded.

[0107] In fact, according to this embodiment, the comparator module 403 is further configured to generate each logic stop signal STP based on the comparison result performed.

[0108] The comparator module 403 can be based on different standards, for example, TTL standard, NIM standard, CMOS standard, ECL standard, etc.

[0109] The comparator module 403 is downstream of the level adapter module 402.

[0110] In this embodiment, the device 200 further includes a digital-to-analog converter module 404 operably connected to the comparator module 403 and configured to provide a tunable threshold to the comparator module 403.

[0111] The digital-to-analog converter module 404 is controlled by the control module 205.

[0112] According to Figure 4 In another illustrated embodiment, in combination with any of the foregoing embodiments, the device 200 further includes a Peltier unit 405 operably associated with at least one silicon photomultiplier sensor 201.

[0113] The Peltier unit 405 based on temperature control is configured to control the temperature change of at least one silicon photomultiplier sensor 201.

[0114] In this regard, regardless of whether the randomness is unaffected by temperature variations, studying the relationship between the current pulse frequency and temperature can be beneficial to the end user.

[0115] According to one embodiment, the Peltier unit 405 is embedded in a package that also integrates at least one silicon photomultiplier sensor 201.

[0116] According to Figure 4 the illustrated embodiment, the device 200 includes a buck DC-DC converter module 406 operably connected to the Peltier unit 405.

[0117] The buck DC-DC converter module 406 is configured to bias the Peltier unit 405.

[0118] The buck DC-DC converter module 406 is controlled by the control module 205.

[0119] According to Figure 4 another illustrated embodiment, in combination with any of the foregoing embodiments, the control module 205 includes an output interface 407 configured to provide a random bit sequence S2 to be provided to the end user EU to the end user EU.

[0120] Examples of the output interface 407 can be communication ports of Ethernet type, USB type, or Wi-Fi type.

[0121] According to different embodiments, device integration can be achieved in an efficient, highly customizable, and cost-effective manner.

[0122] In fact, according to an embodiment not shown in the drawings, in combination with any of the foregoing embodiments, at least one silicon photomultiplier sensor 201 can be housed in a transistor outline (TO) package that also embeds the Peltier unit 405 and at the same time makes it easy to shield the at least one silicon photomultiplier sensor 201 from external light.

[0123] Additionally, according to another embodiment not shown in the drawings, in combination with any of the foregoing embodiments, it should be understood that the measurement of time can be performed by different architectures.

[0124] For example, a specific chip can be used to provide the possibility of measuring time from two independent sources at a sub$ / unit cost, with a resolution of the independent sources at the level of 55 ps, time digitization exceeding 24 bits, and the possibility of recording up to five (5) logic stop signals per cycle.

[0125] However, it should be noted that, in terms of power consumption, this embodiment is expected to require at least a few watts, and a typical multi-layer structure on a typical area of 25 to 100 cm 2 is contemplated.

[0126] As long as size and power consumption are important, any of the foregoing embodiments may be combined, and another embodiment not shown in the drawings may be considered.

[0127] According to this embodiment, the time-to-digital converter module 204 can be implemented in a field-programmable gate array (FPGA) chip, and many architectures have proven to be reliable.

[0128] According to this embodiment, at least one silicon photomultiplier sensor 201 is not temperature-controlled, and it can be mounted on a board in a surface mount technology (SMT) package, but temperature feedback for overvoltage stabilization can be provided.

[0129] Such an embodiment can advantageously be suitable for an area of no more than a few cm 2 for power consumption in the range of a few hundred mW.

[0130] According to another embodiment not shown in the drawings, in combination with any of the foregoing embodiments, a low-power miniaturized device can be arranged on an application-specific integrated circuit (ASIC), which benefits from the progress of vertical integration of microelectronic devices and the design of time-to-digital converter modules for LIDAR (light detection and ranging), high-energy physics, and medical applications.

[0131] According to this embodiment, in addition to including at least one silicon photomultiplier sensor 201, the device 200 may further include additional silicon photomultiplier sensors.

[0132] Also referring to Figure 5 , a method 500 for generating a random bit sequence according to the present invention will now be described.

[0133] The method 500 includes a symbolic step start ST.

[0134] The method 500 further includes the step of generating 501 a sequence S1 of endogenous random current pulses by self-amplification of thermally generated charge carriers driven by avalanche ionization in at least one silicon photomultiplier sensor 201, and the at least one silicon photomultiplier sensor 201 is subjected to the thermally generated charge carriers.

[0135] The method 500 further includes the step of receiving 502 the sequence S1 of the endogenous random current pulses by the data processing unit 202 of the device 200 for generating a random bit sequence.

[0136] Method 500 further includes the following steps: determining 503, by data processing unit 202, a random bit sequence S2 to be provided to end user EU based on sequence S1 of the endogenous random current pulses received from the at least one silicon photomultiplier sensor 201.

[0137] Method 500 ends with a symbolic step ED.

[0138] According to one embodiment shown in dashed lines in Figure 5 step 503 of determining further includes the following steps: determining 504, by signal conditioning module 203 of data processing unit 202, a logic stop signal STP for each current pulse in sequence S1 of the endogenous random current pulses received from the at least one silicon photomultiplier sensor 201.

[0139] According to one embodiment shown in dashed lines in Figure 5 step 503 of determining further includes the following steps in combination with the previous embodiment:

[0140] - receiving 505, by time-to-digital converter module 204 of data processing unit 202, a logic start signal STR and each logic stop signal STP determined by signal conditioning module 203;

[0141] - measuring 506, by time-to-digital converter module 204 of data processing unit 202, the time interval between the arrival time of each logic stop signal STP and the arrival time of logic start signal STR.

[0142] According to one embodiment shown in dashed lines in Figure 5 step 503 of determining further includes the following steps in combination with the previous embodiment: providing 507, by control module 205 of data processing unit 202, a logic start signal STR to time-to-digital converter module 204 of data processing unit 202 when device 200 is powered on.

[0143] According to one embodiment shown in dashed lines in Figure 5 step 503 of determining further includes the following steps in combination with the previous embodiment:

[0144] - processing 508, by control module 205, each time interval measured by time-to-digital converter module 204;

[0145] - generating 509, by control module 205, a random bit sequence S2 to be provided to end user EU based on each processed time interval.

[0146] According to one embodiment shown in dashed lines in Figure 5 step 504 of determining includes the following steps:

[0147] - Each current pulse in the sequence S1 of endogenous random current pulses is converted 510 into a corresponding voltage pulse by the current-voltage conversion module 401 of the signal conditioning module 203;

[0148] - The level of each voltage pulse is adjusted 511 by the level adapter module 402 of the signal conditioning module 203;

[0149] - Each voltage pulse is compared 512 with a tunable threshold provided by the digital-to-analog converter module 404 of the device 200 by the comparator module 403 of the signal conditioning module 203;

[0150] - Each logic stop signal STP is generated 513 by the comparator module 403 of the signal conditioning module 203 based on the result of the comparison performed.

[0151] According to an embodiment shown in dashed lines in Figure 5 In combination with any of the foregoing embodiments, method 500 further includes the step of controlling 514 the temperature change of at least one silicon photomultiplier sensor 201 by a Peltier unit 405 of the device 200 operably associated with the at least one silicon photomultiplier sensor 201.

[0152] The device and method according to the present invention have several advantages, namely:

[0153] a) Minimized complexity due to the endogenous nature of the random pulses of quantum nature;

[0154] b) Low power consumption;

[0155] c) Robustness against temperature and power supply variations;

[0156] d) Low cost due to silicon technology;

[0157] e) High efficiency (bit rate / pulse) (without using a whitening algorithm for post-processing);

[0158] f) Cost-effectiveness;

[0159] g) Considerably high bit rate (exceeding 1 Mbps / mm2 sensor);

[0160] h) Scalability.

[0161] Without departing from the scope of the appended claims, those skilled in the art can make modifications and adaptations to the above-described embodiments of the device and method for generating a random bit sequence, or can replace elements with other functionally equivalent elements to meet accidental requirements. Each feature of the features belonging to one possible embodiment can be implemented independently of the other embodiments described.

Claims

1. An apparatus (200) for generating a random bit sequence, comprising: - at least one silicon photomultiplier sensor (201) configured to generate a sequence (S1) of endogenous random current pulses due to self-amplification of thermally generated charge carriers driven by impact ionization, to which the at least one silicon photomultiplier sensor (201) is subjected; - a data processing unit (202) configured to receive the sequence (S1) of endogenous random current pulses and to determine a random bit sequence (S2) to be provided to an end user (EU) based on the sequence (S1) of endogenous random current pulses received from the at least one silicon photomultiplier sensor (201).

2. The device (200) according to claim 1, wherein: The data processing unit (202) comprises a signal conditioning module (203) configured to receive the sequence (S1) of endogenous random current pulses, the signal conditioning module (203) being configured to determine a logical stop signal (STP) for each pulse in the sequence (S1) of endogenous random current pulses received from the at least one silicon photomultiplier sensor (201).

3. The device (200) according to claim 2, wherein: The data processing unit (202) further comprises a time-to-digital converter module (204) configured to receive the logical start signal (STR) and each logical stop signal (STP) determined by the signal conditioning module (203), the time-to-digital converter module (204) further being configured to measure a time interval between an arrival time of each logical stop signal (STP) and an arrival time of the logical start signal (STR).

4. The device (200) according to claim 3, wherein: The data processing unit (202) comprises a control module (205) configured to send the logic start signal (STR) to the time-to-digital converter module (204) when the device (200) is powered on, and the control module (205) is further configured to process each time interval measured by the time-to-digital converter module (204) and generate the random bit sequence (S2) to be provided to the end user (EU) based on each processed time interval.

5. The device (200) according to claim 4 further includes a high-voltage power supply module (206) configured to provide a high-voltage power supply to the at least one silicon photomultiplier sensor (201), and the high-voltage power supply module (206) is also configured to control the dependence of the operating voltage of the at least one silicon photomultiplier sensor (201) on the excess bias voltage relative to the breakdown voltage value.

6. The device (200) according to claim 5, wherein: The control module (205) is configured to control the at least one silicon photomultiplier sensor (201) via the high voltage power supply module (206), and the control module (205) is also configured to control the signal conditioning module (203) and the time-to-digital converter module (204).

7. The device (200) according to claim 4, wherein: The signal conditioning module (203) comprises: - a current-to-voltage conversion module (401) configured to convert each current pulse in the sequence (S1) of endogenous random current pulses into a corresponding voltage pulse; - a level adapter module (402) configured to adapt the level of each voltage pulse; - a comparator module (403) configured to compare each voltage pulse with a tunable threshold provided by a digital-to-analog converter module (404), the digital-to-analog converter module being operably connected to the comparator module (403), the comparator module (403) being further configured to generate each logical stop signal (STP) based on the result of the comparison performed.

8. The device (200) according to the preceding claim 7 further comprises a Peltier cell (405) operably associated with the at least one silicon photomultiplier sensor (201), the Peltier cell (405) based on temperature control being configured to control temperature changes of the at least one silicon photomultiplier sensor (201).

9. The device (200) according to claim 8, wherein: The Peltier cell (405) is embedded in a package that also integrates the at least one silicon photomultiplier sensor (201).

10. The device (200) according to any one of the preceding claims 8 and 9, further comprising a step-down DC-DC converter module (406) operably connected to the Peltier unit (405), the step-down DC-DC converter module (406) being configured to bias the Peltier unit (405), the step-down DC-DC converter module (406) being controlled by the control module (205).

11. The device (200) according to any one of the preceding claims 4 to 9, wherein: The control module (205) comprises an output interface (407) configured to provide the random bit sequence (S2) to be provided to the end user (EU) to the end user (EU).

12. A method (500) for generating a random bit sequence, comprising the steps of: - generating a sequence (S1) of endogenous random current pulses by at least one silicon photomultiplier sensor (201) due to self-amplification of thermally generated charge carriers driven by impact ionization, said at least one silicon photomultiplier sensor being subjected to said thermally generated charge carriers; - receiving the sequence (S1) of endogenous random current pulses by a data processing unit (202) of the device (200) for generating a random bit sequence; - determining, by the data processing unit (202), a random bit sequence (S2) to be provided to an end user (EU) based on the sequence (S1) of endogenous random current pulses received from the at least one silicon photomultiplier sensor (201).

13. The method (500) of claim 12, wherein: The step of determining a random bit sequence (S2) to be provided to an end user (EU) further comprises the step of determining, by a signal conditioning module (203) of the data processing unit (202), a logical stop signal (STP) for each current pulse in the sequence (S1) of endogenous random current pulses received from the at least one silicon photomultiplier sensor (201).

14. The method (500) of claim 13, wherein: The step of determining a random bit sequence (S2) to be provided to an end user (EU) further comprises the following steps: - receiving, by a time-to-digital converter module (204) of said data processing unit (202), a logical start signal (STR) and each logical stop signal (STP) determined by said signal conditioning module (203); - measuring the time interval between the arrival time of each logical stop signal (STP) and the arrival time of the logical start signal (STR) by the time-to-digital converter module (204) of the data processing unit (202).

15. The method (500) of claim 14, wherein: The step of determining a random bit sequence (S2) to be provided to an end user (EU) further comprises the step of providing the logic start signal (STR) to the time-to-digital converter module (204) of the data processing unit (202) by a control module (205) of the data processing unit (202) when the device (200) is powered on.

16. The method (500) of claim 15, wherein: The step of determining a random bit sequence (S2) to be provided to an end user (EU) further comprises the following steps: - processing, by the control module (205), each time interval measured by the time-to-digital converter module (204); - generating, by said control module (205) on a per processed time interval basis, said random bit sequence (S2) to be provided to said end user (EU).

17. The method (500) according to any one of the preceding claims 13 to 16, wherein: The step of determining a logical stop signal (STP) for each current pulse in the sequence (S1) of endogenous random current pulses received from the at least one silicon photomultiplier sensor (201) comprises the following steps: - converting each current pulse in the sequence (S1) of the endogenous random current pulses into a corresponding voltage pulse by the current-voltage conversion module (401) of the signal conditioning module; - adapting the level of each voltage pulse by a level adapter module (402) of the signal conditioning module; - comparing each voltage pulse by a comparator module (403) of the signal conditioning module (203) with a tunable threshold value provided by a digital-to-analog converter module (404) of the device (200); - Each logical stop signal (STP) is generated by said comparator module (403) of said signal conditioning module (203) based on the result of the comparison performed.

18. The method (500) of claim 12, further comprising the steps of: A temperature change of the at least one silicon photomultiplier sensor (201) is controlled by a Peltier cell (405) of the device (200) operatively associated with the at least one silicon photomultiplier sensor (201).

Images