Patent Application
20240089087
This invention relates generally to encryption and, in particular, to encryption systems and methods based upon one-time-pad (OTP) key generation, distribution and use.
The modern Internet based economy is built on the security of public key/private key encryption systems. These systems rely on asymmetric math problems that are hard to solve, but easy to confirm via the correctness of a solution. A common encryption problem is the prime factorization of a very large number. Solving these types of problems requires tediously testing every possible solution until the answer is randomly found. Using conventional computers, cracking this encryption requires incredible amounts of time, expense, and computational power—making it impractical for the typical hacker. The current systems, such as those used for SSL (Secure Socket Layer) in an Internet browser, are very convenient and cheap to maintain and use. Without this technology, it would not be safe to perform financial transactions online, and general user privacy would be compromised.
Decades ago, a new type of computer, a “quantum” computer, was envisioned that would exploit the quirks of quantum mechanics to solve brute force problems almost instantly, by examining all possible solutions simultaneously and settling on the correct answer. About 10 years ago the first proof of concept quantum computers were developed in the lab. Several years later, a company called D-Wave introduced the first commercially available quantum computer.
Early versions were essentially high cost novelties ($10M each) that would allow companies to begin developing algorithms that might become financially viable later. In 2018 D-Wave opened a cloud computing service to the public using its latest generation system. As commercial services are now opening using D-Wave's cloud, there are indications that the technology is crossing the threshold to economic viability. An exponentially faster quantum system will be released by D-Wave shortly. Recently, IBM and Regatti have also released quantum cloud computing using a different technology.
Quantum computers hold great promise for mankind by optimizing the efficiency of complex systems like air traffic routes, aiding the design of pharmaceuticals and machinery, and enabling better artificial intelligence—to name just a few applications. This same power also threatens to destroy the Internet economy by breaching the asymmetric encryption systems used today. The prevailing wisdom is that for now, the length of encryption keys can simply be increased periodically to keep up with the increasing power of quantum computers. This is true to a point. Key length has been increased successfully to compensate for gains in conventional computer capabilities. There is also broad recognition that eventually quantum computers will win out and a new system of encryption will have to be adopted. An additional problem in the interim, is that old messages or transactions could easily be decrypted by next year's quantum technology.
Once quantum computers have reached sufficient power to render asymmetric encryption unsafe, Internet purchases with credit cards will become unsafe. All online banking will stop. Sensitive personal or corporate information could not be transmitted with privacy. Remotely controlled machinery as well as autonomous planes and cars could be hacked into by terrorists. All crypto currencies (such as Bitcoin) rely on asymmetric encryption to maintain their integrity, as do all other current blockchain applications. Much is at stake in replacing current encryption methods in time to prevent a nightmare scenario.
It is well known that a One Time Pad (OTP) encryption system is mathematically impossible to crack, regardless of brute force computing power. Therefore OTP is resistant to cracking by a quantum computer. OTP involves encrypting a message with an XOR logic operation using a random key that is of identical length to the message. Both the sender and receiver use the same key to encrypt and decrypt the message. If the key remains secret, is truly random in nature, and is never reused—then the security of the message is assured. OTP was used extensively in WWII for secure military communications and continues to be used for highly sensitive government and corporate data. Various software programs can be found on the Internet enabling OTP communication between two users. These applications typically require users to physically meet to share a key.
While fully secure, OTP presents practical issues. First of all, keys must be the same length as the message and never reused. This necessitates a large supply of encryption keys that must be securely delivered beforehand between senders and receivers. A different set of keys must exist for all combinations of senders and receivers. Additionally, the keys must be truly random—so pseudo-random number generators cannot be used safely. For these reasons, OTP has been pushed aside for public key/private key encryption systems.
It is recognized in the trade literature that adopting a “star network” would overcome the problem of connecting every possible combination of sender and receiver in a network. A communication could be sent to a central encryption hub, decrypted, and then re-encrypted and sent to the final recipient. This allows every user of the network to maintain encryption keys with the hub only, and still send secure communications to anyone in the network—without any prior contact. This type of system already exists throughout the public telecom network and Internet, as different transmission protocols encrypt and decrypt data to deliver it between disconnected senders and receivers.
In fact, methodologies for using a hub to administer an OTP network have been described and patented. Till now, these systems have contained two key issues. Systems where the hub brokers the messages, rely on the security of the hub—physically, electronically, and in terms of employee integrity. As the volume of financial and other data moving through such a hub increases, the payoff for a bad actor to compromise it becomes enormous. A hub handling financial transactions for millions of users could have billions or even trillions of dollars of sensitive data at stake—in a single location. A chain is only as strong as its weakest link, and the practical difficulty of physically securing such a location would be enormous. Additionally, users of the network would by nature be placing their trust in the administrators of the hub, and would be giving up privacy in this regard.
Another critical problem in described hub systems is that encryption keys and or messages are deleted to minimize long term risk of physical breach. Such a protocol could be compromised and replaced by a bad actor, without the knowledge of users or even the administrators. While the deletion protocol was successfully carried out, the system would present a threat to the national security of the United States or other countries. Terrorists would be able to use such a communications network to communicate securely without interception by the military or law enforcement. Since 9/11, western nations have recognized the need to balance personal privacy and protection from hackers, with an ability for governments to track terrorists in extreme situations. It is unlikely that western governments would permit wide scale use of a system that does not maintain that balance.
This invention, referred to herein as “Quantum Lock™,” resides in systems and methodologies for using quantum resistant one-time-pad (OTP) encryption in ways that resolve the critical problems outlined above. The system preserves the required elements of an unbreakable OTP system (under Claude Shannon's math proof), which some described systems in the prior art do not. Additionally, the invention uses a secure network to reconstitute blockchain systems without the use of asymmetric encryption. Extensions of this block chain system are described which enable an entirely new set of applications for protecting privacy, sharing information, performing validations and analysis of data, and creating system actions that are constrained by complex data algorithms.
In accordance with the invention, random One Time Pad (OTP) keys are generated using a true random number generator and placed on serialized USB flash drives, discs, hard drives, or another data storage device. The devices may be sold in stores, automated kiosks, or purchased through mail order. Standard packaging techniques can be used to encapsulate the memory devices to indicate physical tampering. The automated kiosks would allow recycling and replacement of old OTP key storage devices. A universal or serialized physical or electronic marking would allow the kiosks to identify old devices and accept them—providing a recycling credit towards the purchase of new keys. Alternately, a credit system could be established by which a human confirmed validity of returned devices. A new device already fully loaded could be obtained instantly, as in a first-time purchase. Alternately, the kiosk could load new keys onto the returned device—with a passcode system allowing the user to return later to pick it up after the download is complete. An alternate feature allows downloading onto a reused device, while leaving existing OTP keys (or unrelated files) intact. Users could purchase different volumes of OTP keys. Another embodiment of the system of recycling existing devices would incorporate some human involvement in the download process using a standard computer or custom device.
For the system to be fully secure under Claude Shannon's proof, the encryption keys must be completely random. The pseudo random generators (utilizing a math algorithm) found in many commercial computers, are not truly random. A preferred method for obtaining random numbers is through a physical apparatus that exploits quantum randomness. An example of this is the Quantis device made by IDQ. The servers generating ciphers for distribution could use the Quantis or a similar product to achieve true randomness.
The key storage devices (KSD) would be uploaded with OTP keys and identifying marks (showing it originated from Quantum Lock) at one or more centralized facilities. These facilities would either generate streams of truly random sequences or obtain them from other facilities with this capability. Transmission of OTP keys between facilities would occur via physical movement of memory devices, or via electronic links using OTP encryption. Facilities might periodically physically transfer very large memory devices containing OTP keys for the purpose of establishing a link between facilities. Data could then be transmitted between facilities over Internet or telecom networks later, without the use of a physical delivery service. KSDs would then be physically transported to retail outlets, kiosks, or delivered directly to recipients. Another embodiment would be for kiosks or retail locations to have the capabilities to generate the OTP keys on site. In this case, copies of the keys would be physically transported (or sent back via OTP encryption) to a central facility. Another embodiment would permit download of keys onto a storage device not acquired from Quantum Lock; in this scenario no recycling would be permitted. Still another option for the kiosks would be to enable wireless download of keys onto a cell phone, computer, or other electronic device.
Software to manage the Quantum Lock system would be distributed on the KSDs, on a separate storage medium, or downloaded directly over the Internet or telecom network. The software would manage the OTP keys, carry out encryption and decryption, interface with other software programs on the user's computer, and assist the user with long term indexing, deletion, and archive functions of already used keys sequences. The software could be publicly available or might be restricted to subscribers of the Quantum Lock system. Alternately, the software could be customized for each user. The software and or version updates might themselves be distributed only via physical transport or OTP encryption to prevent public examination. Copies of distributed OTP keys are stored in a server hub and associated with a KSD and/or a user account. The hub would function as an exchanger of keys between senders and recipients, and as a long-term storage facility of keys. Multiple hubs might be employed to reduce communication latency over large geographic areas. Multiple hubs might also store redundant copies of keys to protect against physical damage to a single facility (by fire, flood, etc.). Redundancy could also be used to balance load the volumes of data being transmitted by a single facility, and to keep the network functioning in the event of a temporary failure at a single facility due to power loss or another technical failure. Redundancy of keys might exist on multiple servers at a single hub facility, as well as non-hub facilities that do not communicate with users and are used only for data storage. Users might have the ability to communicate with all hubs, a single hub, or a subset of hubs.
Facilities storing keys would require strong physical and electronic protections. Internet or telecom access to a hub would be necessary for functioning. Servers containing copies of keys might be separated from those with access to the outside world. They could communicate over a port governed by software creating a restriction around the port that allows transmission and retrieval of keys only—and provides no access to the server operating system. An embodiment of this would physically “gap” the servers, and a memory device would physically transfer communication between the servers.
Servers would store OTP keys for varying lengths of time based on the desires of specific users, universal system settings, or based on government mandates. Expended keys might be moved after use, and redundancies could be increased or decreased over time. Users would be able to request copies of old keys (such as in the event of losing a KSD at their home or business). This would allow them to view an encrypted file in their possession that they could not access because of losing the key. Copies of old keys could be sent out on a new KSD, uploaded onto an old KSD with adequate free memory space, or transmitted via OTP over the Internet or telecom network. Requests for old keys could occur electronically with security measures such as passcodes, questions, identifying data such as social security numbers, or biometric security. Alternately, a human could be involved in fulfilling such requests and verifying identity in the same way a bank teller might. Additional precautions in releasing old copies of keys might include sending KSDs only to the address on file of the user. In the event a valid judicial warrant was presented for a governmental agency to acquire a key, the requested copies could be delivered through similar means and with similar verification of validity.
The address of the recipients could be an email address or another type of user identifier that exists apart from the Quantum Lock system or is assigned by the system. The messages or data transmitted in the Quantum Lock system could be sent via email, FTP, HTTP, another Internet protocol; or via a non-Internet telecom network, or via direct landline or wireless transmission using sound, photons, electromagnetic waves, or particles; or via physical transport on a storage medium.
The software may manage multiple KSDs to create a master cipher for a user that constitutes the sum of all ciphers. This would allow the end of the key on a KSD to be attached to the beginning of the key on another KSD for efficiency. Keys from multiple KSDs might also be combined and copied onto other (possibly larger) storage devices. In this embodiment, a header at the beginning of a master file would denote the sizes and serials of the sub cipher keys compiled together, as opposed to any marker within the cipher. Should a marking sequence be used inside a master cipher to denote the beginning of admin data, this would itself violate the integrity of the OTP system as that sequence would have to be purged from random keys generated.
The software and hub would also use industry standard mechanisms to achieve load balancing or rerouting between hubs. Error protocols would be created to account for transmissions that are not completed, so that cipher markers are synced, and messages reprocessed when necessary. Admin data stored locally by the user software could also be backed up for redundancy at Quantum Lock hubs. Back up of encrypted messages might be stored at a cloud storage facility unassociated with the hubs—to protect the separation that is necessary to the system's integrity.
The software could encrypt messages with the appropriate cipher and add a file extension to the name such as “.qlock”. A header protocol would allocate a fixed length at the beginning of each file for admin data such as the message identifier, sender, and recipient. The software would match this data within a message to an index of ciphers that is compiled over time through communications with the hub network.
Encrypted files could easily be sent as attachments to email. Over time, interfaces with other software programs (and electronic hardware devices) could be developed. For example, all Internet browsers contain capabilities for asymmetric encryption (HTTPS). A so-called plugin or addon seamlessly replacing the current encryption with an interface to the Quantum Lock system would allow financial transactions to continue on the Internet, unimpeded by the advance of quantum computers. Entirely new communication software using Quantum Lock could also be developed. Hardware devices using Quantum Lock could also be developed that sit between an electronic device and the public Internet or telecom system (such as a copy or fax machine, computer, router, drone, or another machine or device controlled remotely). Such a device might encrypt all data coming in or out of the protected device, using the Quantum Lock system; and route all data directly through Quantum Lock hubs. This could be done without any knowledge of the proprietary workings of software or hardware on the other side of the barrier. For example, a Quantum Lock routing device could be attached to a military drone or an autonomous vehicle linking it back to the authorized operators. This could avert a major loss of life or property by preventing a breach by a terrorist or criminal.
One application of the Quantum Lock system is to create fully secure cloud storage or backup systems. This is easily accomplished by having two or more independent backup systems, run by different companies. The Quantum Lock system would provide a cipher to encrypt the data being sent to the backup system. One service would store the cipher, while the other would store the encrypted data. Neither would see the raw data, maintaining the user's privacy. When the data was recovered, the cipher and encrypted data would simply be combined. Transmission of information to the storage services would be through Quantum Lock, ensuring security and privacy in transit. Users could in theory “daisy chain” multiple storage systems and ciphers together for adding security. In order to obtain the unencrypted data, a bad actor would require multiple ciphers. For example, the raw data could be encrypted five times via XOR OTP, and each of the ciphers stored in a separate backup service.
An additional service that flows naturally from the Quantum Lock system is a one-time passcode service. Some banks physically give their clients a device that contains a series of one-time passcodes to increase the security of transactions. Some systems store a series of codes used in sequence, while others use more complex schemes such as codes based on a time algorithm. One-time passcodes add security because interception of a passcode is useless, as it immediately expires. A criminal would need to obtain the source of the passcodes to carry out a fraudulent transaction.
It is tremendously expensive and inconvenient for one time passcodes to be used currently, as there is no universal exchange. A device must be physically transferred for every relationship requiring the added security. It is the same problem one encounters in establishing a point to point One Time Pad system. Quantum Lock overcomes this by acting as a exchanger of ciphers, so that cipher need only move between Quantum Lock and each user—not between every possible combination of users.
A one time passcode system can be created easily by one party sending the other party the block of one time codes via Quantum Lock. The codes could be created by using a portion of a Quantum Lock cipher, since these are already random, or by another means. This would be vastly more efficient and cost-effective than what is currently happening in the financial world. The ease of use would open up one time passcodes for applications other than larger bank transfers, including general website logins.
The Quantum Lock system allows a message to be sent with full security over the public Internet without users physically or electronically having to directly exchange the encryption key. This is vastly more efficient than having to maintain a different cipher for every possible message recipient. Additionally, the server never handles the message, so privacy and security related to the hub or network administrators are not a concern.
Hackers could not brute force decrypt any intercepted message that is “Quantum Locked”, even with a quantum computer. They would have to obtain a copy of the encrypted message from the public Internet as well as the key from the hub facility. This would provide extraordinary privacy and security safeguards to private citizens and businesses. The Quantum Lock system would also protect against hacking by rogue foreign nations. However, governments typically can wiretap their own telecom and Internet systems. Therefore, governments could obtain an encrypted message via wiretap (as the hub would not have it) and then decrypt it by presenting a valid warrant to the hub. This unique system protects the balance of personal privacy with national security, while preventing an economic upheaval or collapse from the advent of quantum computers.
A unique aspect of the Quantum Lock system enables exceptionally large data transfers with perfect security, without the prior transfer of an equally large cipher. Einstein theorized that information cannot travel in our universe faster than the speed of light in a vacuum, denoted by the constant c. More than a century of experimental and theoretical work in physics has upheld this law as one of the most basic principles of science.
Imagine a signal sent between two parties, Alice and Bob, using light in a vacuum. An eavesdropper intercepts the signal, records it, and then retransmits an identical light signal so that neither party knows their privacy was compromised. Improved engineering and technology allow for incrementally faster interception and retransmission. However, the laws of physics require that any interaction of this nature will introduce a time delay in the reception of the signal.
If Alice and Bob have synchronized clocks and agree on a predetermined time for a message to be sent—then they can know if an interception has occurred. The precision of their clocks must be such that they can detect discrepancies smaller than the latency introduced by the eavesdropper's equipment.
If the time of transmission is sent in the signal itself, it could simply be tampered with to adjust for the latency. Alice and Bob can exchange a Quantum Locked message to eliminate this problem. The message could be very short, for example simply containing the time when a message is sent and its length, and a description if necessary, of the time interval between the data pulses. This message might be sent ahead of time, simultaneous to the other transmission, or afterwards. The key to security is that Alice and Bob exchange a Shannon secure message separate from the body of the text which would be either unencrypted or encrypted using non-Shannon secure encryption. If they send an insecure message with the time of transmission, then the message can be tampered with along with the transmission itself.
Another embodiment would create a protocol for Quantum Locked header data to be included in the transmission with the time of transmission, and any other necessary information to describe the data transfer. A tampering party would not be able to adjust the time in the header because they could not decipher it.
It should be noted that most communication systems do not operate at exactly the speed of light. Many operate close to the speed of light as transmissions occur through electric wires, or mediums such as air, or inside a fiber optic cable. In such an instance, the total time in transit between Alice and Bob must be calculated based on this adjustment in speed. If the time in transit is too great relative to the speed of light in a vacuum on a direct line of sight, then in theory an eavesdropper could tap a line of communication discreetly by creating an alternate faster route. This would be complicated, involving the early interception and destruction of a transmission, and reintroduction at a point close to reception. In the rare situation where this was a plausible concern, relay stations could be created along a communications route (such as a trans-Atlantic cable) to either validate data midway, or add a Quantum Locked key to each transmitted message conveying that the data had passed that point.
One critical component of this system is the requirement to measure the signal, or number of transmitting particles, at both the sender and receiver. For example, suppose that a sender transmits a signal that has exactly 20 photons in each pulse. In theory, an eavesdropper could absorb 10 of the photons without cloning them. 10 photons would arrive at the receiver without any latency. To ensure privacy of the message, detection equipment must measure very precisely the number of photons or electrons (or other means of transmission). This engineering challenge is significant, but perhaps much smaller than the challenges presented by attempts to create a “Quantum Internet” which utilizes quantum superposition across vast distances and many hub interactions to achieve privacy.
Most modern communication systems are not point to point but involve complex system of hubs. As such, each segment of transmission between a hub could be validated to ensure there has been no compromise of privacy. Given the constancy of hub positions in most telecom systems, calculating a transmission timetable would be easy. Comparison of time in transit must be calculated based on the exact time sending and receiving hardware exchange data. Many hubs buffer data and relay it based on prioritization. These internal delays due to buffering and analysis of signals would generally not be used for validation of data as times would often be inconsistent.
In some special cases, the effects of general and special relativity might need to be accounted for. For example, suppose a spacecraft had Quantum Locked data denoting a secure transmission to occur at a specific time one month into its voyage. Significant acceleration or the presence of a strong gravitational field would slow the clocks down on the vessel relative to the clocks at mission control on Earth. These effects might be small, but extreme accuracy in the clocks is needed to surpass the speed of modern electronics.
An alternate embodiment involves transmissions from one sender to a plurality of receivers. For additional security, an alternate embodiment could involve multiple transmission paths which are each validated by measurement of the time in transit—in order to confirm a single message.
In many cases, it may be unacceptable for the data in the transmission to be intercepted, even if the receiver will be aware of the interception immediately. To eliminate this problem, a One Time Pad cipher can be sent in the transmission—without any of the actual information that needs to be conveyed. If the receiver confirms the cipher arrived without interception, then a Shannon secure message (via Quantum Lock or another means) can be returned to the sender declaring this. The sender can then release an encrypted message over any public channel at any speed, using the previously transmitted cipher for encryption. It will not matter if this message is intercepted. When it arrives, the receiver can decrypt it with the cipher knowing that no one has viewed it.
Potential embodiments of the invention involve transmission through different types of mediums including gases, liquids, and solid matter. Transmission may be via electromagnetic waves, electric or magnetic fields, acoustic waves, electricity, free electrons, neutrinos, or other particles. Transmission may involve any speed of propagation, and any wave frequency.
The invention might use the Quantum Lock system to relay Shannon Secure data regarding transmissions, or a different OTP or post-quantum encryption system. The format of data transmission may vary, as well as the format and encryption of the data contained in the transmitted messages.
Hardware could be developed to attach to existing communication systems to facilitate this new system of data validation. The hardware would require significant precision in measuring the time that transmissions are sent and received, and the ability to interface with Quantum Lock or a similar system. This invention permits the fully secure transmission of vast amounts of data with only a small amount of cipher being physically transported beforehand.
As mentioned earlier, the demise of asymmetric encryption will make current blockchain applications and cryptocurrencies essentially non-functional. A novel approach to reconstitute blockchains is possible building on the Quantum Lock encryption just described. The system and method described herein represent a complete revision, not merely a swapping of encryption schemes. However, it achieves similar objectives without the possibility of hacking by quantum computers. An immediate advantage over current systems is that the computational power required for functioning is minimal. Bitcoin and other crypto currencies, while still representing a very small portion of the world's money supply—already use more electricity than a small nation. This represents an inherent economic inefficiency and would eventually create environmental concerns as well.
Blockchain technology is revolutionary in that it creates a distributed storage, processing, and verification system—that is “trust minimized”. Rather than relying on a single institution to ratify a transaction or confirm information, blockchains provide public ledgers or histories. Multiple, possibly random, parties work together to add to this ledger by creating blocks of newly verified information that is built on and linked to already trusted blocks in the public ledger. Error checking mechanisms exist to rescind bad blocks or bad chains that are later found to be inconsistent with the public knowledge. All trust in current blockchains is placed in the security of the asymmetric encryption systems built into the blocks—the blocks are typically tied together through a mathematical algorithm called hashing. Due to this, the advent of powerful quantum computers might make trust minimized blockchains, the least trusted mechanism for electronic or remote transactions.
In a “Quantum Locked” blockchain, blocks of information are “chained” together using the Quantum Lock OTP encryption system rather than public/private key asymmetric encryption. While a public ledger might be generally available unencrypted, using the blockchain system for verification or creation of blocks would require being connected through the Quantum Lock hub system. A block contains new information that is related to a prior block (or chain of blocks); except for a “genesis” block. An example would be an automobile title transfer. The genesis block in this example would be created when the car is manufactured and designate the factory as the owner. As the car changed hands, new blocks would be added to that chain showing the current and previous owners, and possibly many other pieces of information such as repairs and accidents. A “terminus” block would be the final block in a chain (perhaps when a vehicle is crushed); no further blocks could be added.
Each block would have required pieces of information and might have optional fields as well. There might be formatting requirements for the fields based on the application. When a sponsoring association establishes a Quantum Locked Blockchain Network (QLBCN), it will have options for establishing these parameters. The parameters could be changed over time, and so a block might have a field identifying the protocol version that was used during creation. At the bare minimum, a block would contain a network id, a block id, timestamp, a pointer to the prior block, pointers to the “VN Nodes” (VNs) of the old block, pointers to the current VNs, pointers to the current owner of the block, pointers to the next block (when established), pointers to the next block's VNs, a field noting genesis or terminus blocks, and one or more data fields. A VN is an entity responsible for storing and confirming data for a specific block, using encryption keys. This is an important concept and will be described in more detail later.
It is understood that some fields could be combined using a formatting protocol. Additionally, or alternatively some of the field data for a block could be stored in one or more indexes and referenced with the block id. It is also understood genesis or terminus blocks would naturally not reference prior or future blocks based on their status.
An example of a simple block for our automobile title network would be as follows:
Note that the IDs in the fields can be alphanumeric or descriptive text or both—however they must serve as unique identifiers. Note also, that some QLBCN's could permit branching or merging of chains. In such cases, the additional prior or future block data would be referenced. For example, a blockchain for land titles might need to account for parcels of land being split apart or consolidated.
Associations creating QLBCN's would have an option to create tiers of VNs with different levels of permission. They would also set the criteria for becoming a VN at a certain level and exiting or being removed as a VN. Examples of VN levels might include the rights to create genesis or terminus blocks, verification of existing blocks only, creation of new blocks, and archival storage of blocks. In some instances, a VN might need to be a law firm or have some other accreditation, or perhaps some verification functions could be done by any random public entity. A system of compensation would also be chosen for VNs completing different functions (such as a penny for each block verification performed). Archival parameters could be used to limit the size of the active public ledger or optimize processing functions. For example, active chains might be limited to 10 blocks with older blocks referenced in the archive ledger(s).
Each block would have at least two categories (but typically three or more) of VN “chains” responsible for ratifying the legitimacy of the block data. One or more chains would be categorized as VNs chosen by (and representing the association or public interest). A set of chains would be categorized as VNs representing each party with involvement in a transaction or with the data in a block. So title transfers for currency or physical assets (like cars or land) would typically have two sets of VN chains—one for the buyer and one for the seller. A blockchain ratifying multi-lateral peace treaties might have many categories of VNs, representing every interested party.
Security of a VN chain could be increased by creating secret keys exchanged between VNs and other VNs, possibly every other VN in a chain, as well as VNs and block owners for previous linked blocks. This would be in addition to the keys held between VNs and the owner of the block a VN is assigned to. The cross linkages between VNs would make it more difficult to impersonate a VN, as the bad actor would require many different secret keys held between many parties. If communication is achieved via the Quantum Lock system, additional communication keys are required compounding the difficulty of cheating the system.
One method for linking every VN with every other VN in a chain would involve the following process. The first VN in the chain would encrypt the block data with its secret key held with the block owner. It would then encrypt the data with the secret keys held with each of the other VNs in the chain. The encrypted data would be sent on to the next VN in the chain who would repeat the process with their corresponding OTP keys. The last VN would send the result on to the sender of the query.
Note that the sender of the query will still use the same process of decrypting the returned data with the composite XOR key supplied by the block owner. All the linkages between VNs will simply cancel out if all is well. If a VN was being impersonated, and did not have one of the linkage keys, then it would corrupt the data. This cancellation of the VN cross-linkage keys happens because performing a XOR function twice has no net effect.
In a network configuration where there are no VN chains, each VN could still hold cross linking keys with all or a portion of the other VNs. In this scenario, the VN could send their cross-linking keys directly to the query sender for matching across the body of VNs involved in the authentication.
The nature of the OTP system used in Quantum Lock ensures that the encryption keys held by VNs can not be cracked by brute force or quantum computation. However, the novel system of VN chains representing the various interested parties, serves to prevent bribery, corruption, or collusion in falsifying blocks. While the encryption algorithm will be described shortly, suffice it to say now that an adequate level of positive verification must be obtained from all categories of VNs before a block can be created or verified later. Each chain of VNs will work in sequence to confirm or deny a request. The entire chain of VNs must provide a perfect affirmation, or the chains “vote” will be considered a denial. By allowing multiple chains of VNs in each representation category—not all chains would need to affirm a block in a given request. This would allow for a broken chain because of technical malfunction or slow response by a specific VN.
The QLBCN would have parameters set on how long the VN chains would be, and the thresholds for system verification in terms of the number or percentage of chains that must respond to ratify a given request. Error protocols would also be established for dealing with situations where both positive and negative votes are cast, and for replacing corrupt chains. The longer the VN chains are and the more required to respond to a request, the more secure the network. This added security also increases the potential processing time for request verification. A network can be designed to create the balance of factors that is best for the application. Note that different levels of security might be required for different sized transactions or blocks of varying importance. For example, a block recording the purchase of a cup of coffee might only have three VNs per chain, while a billion-dollar stock exchange might require 100. It will be demonstrated later, after the encryption protocol is explained fully, how a carefully designed QLBCN would make it fantastically difficult to forge false blocks—achieving a truly distributed, trust minimized network.
The public information of the block would contain only the first and last VN in each chain. VNs connecting the endpoints would only be known entirely to the owner of the block. This adds a layer of security by hindering bribery of VNs. A VN would also be known to the VN just prior to them on the chain. Each VN would maintain a secret OTP key with the owner of the block, identical in length to the block. When a system user wants to confirm ownership of a block, it will send the block to the starting VN in each chain. The VN then encrypts the block with the secret key and transmits it to the next VN. The next VN performs another XOR operation on the data with their secret OTP key, and sends the data on. The final VN in the chain, knowing it is the endpoint, sends on the encrypted block back to the user making the query. If the chain has 11 VNs in the chain, the block will be encrypted by XOR operation 11 times. The owner of the block will be asked to release the composite XOR key for the entire chain, back to the user. The user will then decrypt the garbled data from the endpoint of the VN chain with the composite key from the owner. If it matches the block—then this indicates that the claimed owner truly owns the block as verified by that chain. Each chain will be queried in similar fashion, and based on the rules of the network, a threshold of agreement can be reached.
An immediate issue that presents itself is that, once a single query has been run; the composite OTP key for each VN chain can be acquired. This would allow a bad actor to impersonate the owner of the block. To overcome this, the secret key held by each VN and the owner; can be replaced for each query. As a practical matter, a set of keys could be put in place ahead of time to minimize delays in execution later. For example, when a block is created 100 keys could be put in place between the owner and each VN. A counter can be incremented with each query and owners and VNs in each chain could move to the next key in unison. Once the cache of keys is running low, then another block of keys can be transferred.
Another embodiment of the system would utilize hidden VN keys that are lesser or greater in length than the block length. Another would specifically use lesser lengths which add up to the full length of the block, and each VN would encrypt a portion of the block as it is passed along. In another embodiment the keys would be added together or used to encrypt the prior key in the VN, without performing an action on the block data (the key coming from the end of the VN chain would simply be compared with the secret key of the block owner).
It is assumed by default that all key transmission and all communication in the QLBCN is done through the Quantum Lock system to ensure security of the data. Another embodiment of the system would permit the use of an alternate OTP system that adheres to or falls within Shannon Clark's proof. Another embodiment would permit the use of transmission via quantum entangled particles such that tampering would be impossible or detectable (small scale systems have been developed to date). Alternately, a QLBCN could utilize another form of “postquantum” cryptography that was mathematically proven to be unbreakable (or difficult enough to satisfy the security needs of an application) by quantum or brute force computation.
While the QLBCN can not be breached by malicious decryption, in theory various nodes in the distributed network could be physically breached or bribed. The sheer number of VN nodes involved in a well-constructed network would minimize this risk. One potential point of attack would be to bribe the published end point in a VN chain, while impersonating the block owner. The end point VN could replace the data from earlier in the chain with data that would match a fake hidden key being presented by the bad actor. One solution to addressing this is to have rating systems for the trustworthiness of entities operating a VN node. For example, a well established law firm with a security team might be given a higher rating than an individual that has just begun using the system and verifying transactions. Published VNs at the beginning and end points might be restricted to a smaller pool of highly trusted entities.
An alternate method to address the above concern, would be to publish only the beginning VN in a chain. A user submitting a query would simply receive the response back from the end of the chain and would not know who to expect it from. This eliminates any published VN from returning a fraudulent data set to the user. However, by not knowing the endpoint of the chain; it is conceivable that the beginning VN could collude with a fake VN (one not assigned to a block) and have them send a terminating chain data set to the user that matched a fraudulent key presented by a fake owner.
If this type of collusion and fraud were a concern, a checksum system could be used to verify the identities of the VNs in a chain returning hidden keys or block data encrypted with the keys. In this scenario, each VN in the chain would communicate back directly with the user making the query. The user would match the checksum in the public block data with the VN identities to ensure that each VN was legitimately assigned to the chain for that block. This would prevent a bad actor from compromising, for example, 15 random VNs (out of 10,000 attempts)—and then presenting them as a fake VN chain. The system identifiers for the VNs would not match the checksum unless the bad actor was incredibly lucky. Note that if the VN nodes in the network are all publicly known, then in some cases a brute force or quantum computation on the check sum could reverse engineer the correct combination for a block. Therefore, in this embodiment the checksum algorithm relative to the methodology for assigning identities should allow for numerous potential combinations within a given checksum but the set of those combinations should be a very small fraction of all possible combinations of VNs. This preserves the secrecy of the VNs (so they cannot be targeted for compromise) and prevents presentation of a random set of compromised VNs as a chain. It should be noted that this variation would logically require the VNs to be changed after each query, and the checksum changed in the public data of the block. Checksums might also be used to describe the content of a data block, or of keys being transferred—as part of standard data transmission protocols.
In many of the variations described above, it may be beneficial to periodically alter the VNs assigned to a block. This could involve changing only the secret VNs, or might also include the published VNs. Reconfiguration might be triggered by a query on a block, or could be done periodically system wide based on a time interval. The ledger of public blocks might provide for changing certain aspects over time, such as a checksum or a VN list, but most likely making these changes would require creation of a new dependent block. A QLBCN might be configured such that new blocks are continuously being generated at rapid time intervals—for example every second. If a block had no change of ownership or no new data, it would simply spawn a cloned dependent block with a new set of VNs. Such a system is feasible given the modern speed of communication and computation. The difficulty of compromising the integrity of a vast network of continually changing VN nodes, protecting a block, would be significant.
Maintaining the secrecy of some of the VNs would make it very difficult to compromise all the nodes protecting a block—because they would be unknown. An interesting application of secret VNs is that it would be possible in some variations for them to store data without knowing the content (if both the block data and the block id itself were encrypted). This might have applications of its own. As a practical matter, technical issues such as data corruption or communication latency might be greater issues than wide spread fraud or collusion. A trusted VN's long term business prospects would be placed in jeopardy by attempting to cheat the system, in the same way that various licensed professionals would lose their operating license for perpetrating a single violating scheme. That said, another embodiment of a QLBCN would be to dispense with all secret VNs and publish the identities in the public block. Another configuration would be to dispense with any categories of VNs (representing different interests in a block), or to dispense with any chains of VNs. Conceivably, one could obtain a trust minimized, distributed network by simply having a very large number of VNs assigned to a block; and creating appropriate rules for assessing their responses to a verification query. For example, once most assigned VNs confirm a data set, with no negative responses, then the data is considered validated.
It is intrinsic to blockchain systems, that the older a block is in the public ledger uncontested, the more trustworthy. A QLBCN would operate on this same principle. New blocks can be linked or “entangled” with old blocks by having the owner of a block exchange hidden keys with the VNs and or owners of previous blocks in the chain. This would prevent someone from presenting a completely fake block (not showing up in the ledger yet) and gaining validation through a set of fake VNs. A thorough query would validate the keys to the purported VNs of the new block—but also those in the old parent block which are listed in the public ledger. Different levels of investigation might be carried out based on the importance of the block and the intended interaction. For example, negotiations to purchase an item might begin with a simple query of one or two levels of the blockchain, whereas consummation of a new block showing ownership transfer might require full validation back several levels. Each network could create their own specifications for this. Naturally, in the event an entire chain of blocks was presented contrary to what was propagated (or two diverging versions of the public ledger were propagated) then a pre-established system of forensic investigation would need to be carried out to correct the ledger.
Choice of VNs is an important concept in a QLBCN. If VN categories are being used to represent the interests of different parties, then those VNs might be selected by that party. These might be chosen randomly within a grouping of VNs that the party trusted. Generally, the association building the network would want to assign random VNs to represent the public interest. It is also possible that all VNs for a block would be completely random. The randomness adds a significant amount of security to the system. Ensuring that true randomness is in play would be important in ensuring that an elaborate system of fraud was not created by compromising the verification system itself. Many mechanisms could alleviate this concern. Third party entities could exist to facilitate VN assignments if all parties acknowledged their independence. Each interested party could present a series of keys or numbers that are then added together to appoint VNs (such that no party could influence the outcome). The parties could agree to accept a publicly streamed data value at a specific future point (such as a stock ticker or broadcasted stream from a random number generator). This problem is certainly not unique to QLBCN, but ensuring that it is addressed in network construction is important.
Over time, VNs might withdraw from a network or be removed. As described above, some embodiments of a QLBCN might call for changing VNs of a specific block over time. This presents a problem in that public data for a block must necessarily change. As to VNs simply leaving a network, this could be resolved by making a VN identity (or address) permanent. When a VN left, their identity would be assumed by a new entity. A more robust solution would be for blocks in the public ledger to contain a multi-dimensional array showing the current VNs, as well as a full history. This might include the timestamps of changeovers and detail changes to the VN configuration (such as changing the length of a VN chain). Another embodiment would be to store the VN data in a network of redundant centralized hubs that are recognized for independence. Such a hub might use an internal protocol like the one described above for Quantum Lock encryption hubs, whereby it would not be possible to compromise the operating system via the Internet. Alternately, the VN data could be stored in a distributed public ledger. Separation of this ledger from the regular blockchain ledger might be advantageous as the blocks in the regular ledger would not be changed (only added), while the VN ledger would be continually expanding even without the addition of new blocks.
A QLBCN must establish a mechanism for the dissemination (and correction when necessary) of the public ledger. This process could resemble methods used in current blockchain networks. Current blockchains typically distribute the ledger without any security or use the same type of asymmetric encryption that links the blocks. Ideally, a QLBCN might choose to distribute the public ledger via Quantum Lock or another “post-quantum” cryptography method. This could help prevent temporary localized corruption of the ledger, and additionally allow for private blockchain networks only accessible to a select group.
An additional security option that could be implemented to protect against false blocks (being presented as new blocks), would be to send a broadcast query to every single VN in the network. The real VNs of a block would respond invalidating the credentials of the fake block (others would ignore it). The downside of this mechanism is that for large networks the volume of communication and computation would be significant.
In yet another embodiment of a VN configuration, the VN chains would branch out either through a prescribed pattern or randomly. The branching might decrease latency in the system as a larger number of VNs could be involved with fewer levels. A random branching pattern might give each configuration a unique signature like that of a snowflake. If the unique VN configuration were communicated secretly with parties related to prior blocks, this could serve as a rapid preliminary way of confirming a new blocks validity without checking exchanged keys between parties in the new and old blocks.
Another security feature that could be added to a QLBCN would be for communications by VNs or by all parties to be sent through an anonymizing hub. If a VN configuration involved hiding some VN identities, this would prevent their unmasking through monitoring for communication traffic at the time when a large transaction was known to occur. This would require a bad actor to use surveillance or wiretapping of some type. Another measure that could make it difficult to match communications with known transactions, would be for all VNs to send messages only at specific time intervals.
The security of the public ledger could be enhanced by having designated nodes that help store and distribute the ledger. These nodes could perform a verification query on all new blocks when they are receiving. Smaller nodes or users could place trust in ledger updates or copies that are cross checked against multiple nodes that have performed verification. Ledger nodes should cross validate their copies of the ledger frequently with as many other nodes as possible, perhaps changing who they check with at each validation. This will help uncover corruptions to the ledger rapidly.
If there are multiple owners (or stakeholders) in a block, then keys from each would be held by the VNs. There might be complex rules for getting approval from multiple block owners to perform an action (such as creating a new block transferring ownership of an asset), or simple unanimous approval may be required. For example, transferring a block of cryptocurrency might simply involve verifying the identify of the block owner through the secret keys, whereas issuing new stock in a corporation might require the approvals of most shareholders and certain key officers. Part 3 describes a novel mechanism for organizing complex validation rules in the system.
A QLBCN will grant VNs different rights or roles, as stated earlier. The permissions needed for new block creation will be set as well. There may be different rules for different types of blocks (or blocks representing different types of data). Designated VNs given the rights to create new blocks would act as the intermediaries between the new and old block owners and VNs. They would be tasked with ensuring that the rules of block creation and VN configuration are followed. After a new block is created, and all VNs and secret keys put in place—the block would be pushed out to the public ledger.
Many combinations of the different embodiments described above could be used to construct a QLBCN. The optimal configuration would depend on the nature of the application and operating environment. All configurations contain the following advantages:
QLBCN's can be extended to enable a whole new set of applications that broker privacy and information sharing. One of the greatest problems of the digital age is protecting personal data, while having the ability to easily share it selectively as desired. Companies collect vast amounts of personal data in exchange for free services. A typical American household may have all Internet traffic, physical travel, financial transactions, phone conversations, email, and even personal discussions in their home—recorded and analyzed to create an ever more accurate consumer profile. In theory, this astonishing relinquishment of data is being managed by impersonal devices—but what guarantees are really in place that this is always the case? Along these lines, academic or medical institutions have legitimate needs to tabulate statistical information and must try to do so without compromising individual privacy. Auditing financial transactions or other types of actions is often necessary for managing business relationships or ensuring regulatory compliances.
Four key extensions of QLBCN's permit all the above issues to be addressed. The first is the concept of “Eigenblocks” and “Eigenfields”. An Eigenfield is a field of data in a QLBCN block that performs an algorithmic function on all or part of the general data in a block (for example information on a car title being transferred, not block ids or VN addresses or other technical data needed to maintain the network). The concept is essentially borrowed from the Eigenfunction of quantum mechanics which is used to parse out a specific characteristic of a wave function.
An example of an Eigenfield might be the fuel economy of a vehicle in a car title transfer QLBCN. Another example might be the debt to income ratio of an individual. Still another might be whether someone had one of 13 genes which created a high risk for a specific disease. Note that in each of these examples a vast amount of unrelated data might be contained in the general block (such as a person's entire DNA sequence). Also note that many combinations of related data might exist to produce the value in the Eigenfield (such as the debt to income ratio)—and yet would be unknown through the Eigenfield.
An Eigenfield might be based on an algorithm that is unchanging over time, but it might also be time dependent—such that future blocks in a chain with the same data might produce a different value in an Eigenfield. An Eigenfield might also be based on data external to the current block, and the algorithm might be referenced elsewhere, or data points used in the algorithm (such as what genes cause cancer) might be referenced from elsewhere. Generally, it seems that if an algorithm for an Eigenfield changed, this data should be reflected in a new “cloned” block of the same data so that both operators on the data with timestamps have a historic chronology recorded. However, an alternate embodiment might permit dynamic Eigenfields. Eigenfields would generally be processed through the verification and OTP encryption process of the QLBCN the same as the other fields, but an alternate embodiment might separate them.
An Eigenblock is a block that performs the function of an Eigenfield on an entirely separate block, or a set of blocks. An Eigenblock might be used to aggregate data of a specific type into one place or used to perform statistical or computational analysis on a vast array of blocks. For example, an Eigenblock might list all the cars currently owned by Bob's Car Dealership and show the block pointers for the specific transactions. Another example would be to calculate the entire fuel economy of the vehicles generated by XYZ Car Factory. This would permit rapid regulatory audit with the confidence provided by a blockchain with unbreakable encryption. Another example would be the total number of individuals with a cancer-causing gene, or a probability distribution matrix relating death to the gene.
Since Eigenblocks and fields can associate more than one piece of information, they could be linked together to form complex computational networks such as “neural” networks or artificial intelligence. Both the memory of events and the memory of computation on the events can be locked into the QLBCN.
Eigenblocks might refer to a series of blocks that are static after creation, locked into place in the blockchain with a timestamp. It might also refer to blocks in the network that are not chained together but are updated continuously by a set of VNs for dynamic viewing.
The second important concept in the extension of QLBCN's is that of encrypting data in a block in the public ledger and selectively releasing the key to a party for all or part of the data. An example might be having one's entire DNA sequence in a block, scrambled by an OTP key. If the owner of the block wanted to let a medical facility perform a test for a specific gene, they might release the portion(s) of the full key necessary to evaluate the locations of interest. They would only have access to those areas, without being able to see anything else. They might also be given keys to specific Eigenfields in a block, or to all or part of an Eigenblock.
Permissions given for encrypted block data could be granted to Eigenblocks, rescinded later, or have specific expirations. For example, someone participating in a university statistical study on age and car accidents, might provide select key access to Eigenfields in their QLBCN for auto information. The university might have an Eigenblock aggregating the data which automatically performs the calculations. The university could trust that they are seeing valid data related to their study, and the participant could be assured that the university could not see other information like how much they pay for their cars.
The encryption keys scrambling the public ledger might be changed periodically in the same way that VN keys are replaced. New keys would need to be exchanged with legitimate parties—this could be managed automatically. Changing the keys would guard against theft of keys in the same way that changing your credit card number periodically thwarts future fraud.
An individual or organization could release information very selectively using key management and Eigenblocks and fields. An interesting example would be for a company whose financial transactions are recorded via QLBCN, to establish a system of permissions and Eigenfields—such that an Eigenblock contains the matrix of data for SEC financial disclosures. Each transaction would be categorized and tabulated in real time—with the full confidence level provided by the system.
A third key concept is that of “conditional block execution”. QLBCN's could be constructed such that new block creation has conditionals controlled by Eigenfields and Eigenblocks (in addition to the standard key exchanges by block owners and VNs). For example, imagine that a foundation stored its funds in a QLBCN crypto currency. The foundation's constitution only permits spending funds on relief work in Africa. The creation of a new block effecting funds transfer from the foundation might require the recipient block owner to have a field operator designating it as related to African relief work. Eigenblocks and fields could also be tied to control of objects in the physical world. Conditional block execution allows for the creation of tightly controlled systems, where the rules can not be changed without proper authorization.
The fourth important concept in the extension of QLBCN's is that of an “Anonymizing Computation Block” or ACB. Imagine that a medical R&D company spent billions of dollars developing a complex algorithm that assigned someone a heart attack risk score based on their DNA sequence and other information. They considered their algorithm to be a trade secret. The algorithm requires analysis of significant amounts of personal data that a consumer might not wish to disclose. An ACB could act as a third-party intermediary performing the algorithm on a consumer's data without having full access to either. The output of an ACB could be recorded and timestamped in a QLBCN, allowing the output to be viewed by other parties with credibility.
First the consumer and the medical company would exchange a protocol for scrambling the private data. It might first involve some form of base analysis to organize the data into a format. Then a random protocol for rearranging the storage of the data would be established. For example, the nucleotide at location 11 on the DNA strand might be stored at location 87 now, and the data from location 4 at 33. Every data address would be mapped to a new address via a completely random key. This would scramble the data such that the ACB owner could not realistically know what it was looking at. An additional OTP XOR operation could be performed on top of that potentially, to make it fully indecipherable. The protocol, address mapping, and decryption key (or combination thereof) would be exchanged via Quantum Lock with the medical company using Eigenblocks in a QLBCN.
The medical company's Eigenblocks would take the secret key exchanged and revise its trade secret algorithm to account for the scrambling. This would then be sent to the third-party entity's ACB. The consumer would grant the ACB permissions to the Eigenblock in their QLBCN holding the newly scrambled data. The ACB would be unable to see the individual's data, and not knowing the format of the data—would have difficulty deciphering the medical company's trade secret. The ACB would perform the algorithm on the data, and then return the score to the medical company and or the individual. As such, an ACB is a powerful tool for protecting the privacy of multiple parties whose data must interact to perform a computation.
Note that if too much data is returned by an algorithm given to an ACB—it could compromise the privacy of the user data. For example, an algorithm returning a 500 bit output after operating on a 500 bit input—might simply be unencrypting the data sent to the ACB. If the ACB sends the output to both parties, then the user can see that the algorithm's output is limited to a simple score that protects their overall privacy. In some instances, a party might be concerned that the form of an algorithm is itself is a trade secret—even without knowing the exact nature of the data inputs. This concern can be addressed by creating an ACB chain where an algorithm is broken into multiple pieces and divided up between independently operated ACBs—much like a computer would break up a computation between parallel processors. In this embodiment, a protocol for breaking up the algorithm would be sent to a primary ACB that would direct the interaction of other independent ACBs, who would each receive their portion of the algorithm independently via Quantum Lock.
ACB's have a vast number of applications. For example, targeted ads can be sent to consumers based on their financial, communications, medical, or social media data—without compromising their privacy in any way. Although some anonymization may be used in current schemes, ultimately this highly personal information is stored somewhere, fully accessible to the companies collecting it.
Authentication and voice recognition systems can also use ACB's to protect user privacy. Currently companies that have voice recognition applications store vast amounts of consumer audio, including general conversations in private homes—in order to train their systems. An ACB could accomplish this without these companies having any access to these recordings. Authentication systems based on voice, facial, iris, fingerprints, or DNA could also be based on ACB's, allowing biometric security to be used without infringing on personal privacy.
Insurance and credit scoring could also be done with an ACB to protect consumer privacy, while allowing transparency necessary for efficient markets. Combination with blockchain (particularly a QLBCN) adds even more benefits to this process by locking in histories that cannot be altered.
QLBCN's with ACB technology also create opportunities in the area of Software As A Service (SAAS). Many companies have gravitated towards SAAS because of portability, redundant data backup, and faster or cheaper computation. Providers of SAAS can sometimes reduce piracy and reverse engineering of applications by not having the user host the programming code. The user however gives up the privacy of their data. Systems can be constructed with ACB's wherein the cloud software interacts with user data and inputs protecting the user privacy, as well as the intellectual property of the software provider.
Part 1:
The following is an example of a message sent through the Quantum Lock hub system.
The following example QLBCN block was provided earlier:
The following are examples of simple Eigenfields. Unrelated block fields are omitted for brevity.
This application is the U.S. National Phase application under 35 U.S.C. 371 of International Patent Application No. PCT/US2020046886, filed Aug. 19, 2020, the entire content of which is incorporated herein by reference.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/US2020/046886 | 8/19/2020 | WO |
