BY CHIA-HSIUNG LIU

Following the success of bitcoin, a new research field named "FinTech" has been created, and one of the core technologies in this field is "blockchain." This article will present an overview covering the technical contents of bitcoin and/or blockchain, their non-financial applications, their limitations, suggested solutions, challenges for patent attorneys, and the current status in Taiwan. 
Basic Information of Cryptology
Basically, the algorithms involved in cryptology can be classified into two types, one being algorithms requiring parameters and the other being parameter free algorithms. Such parameters, or "keys," are usually determined by users. If an algorithm is expressed as sRes = Fun (sPara, sInput), the field of sPara is empty for a parameter-free algorithm. However, such field will be used to define a key for algorithms requiring parameters. Any information to be processed by the algorithm Fun will be inputted into the field sInput. The input information for the algorithm may be one or more (original and non-modified) raw data and/or multiple raw data sets, one or more processed data sets, and their combinations. Such processes may be a bit concatenation, a bit reverse, a bit XOR, and/or an algorithm operation. 
Usually, this type of parameter-free algorithm is called a hash function, and MD5 and SHA are the most important algorithms. The output sRes of a hash algorithm is called a hash value. In contrast, an algorithm requiring parameters is called a cryptography algorithm. Algorithms requiring parameters can be further classified into two sub-classes, i.e., one kind being symmetric and the other being asymmetric. In a symmetric algorithm, the keys used for encryption and decryption are identical, and the most important algorithms are DES, 3DES, and AES. In an asymmetric algorithm, the keys used for encryption and decryption are different, one being a private key and the other being a public key, and the most important algorithms are RSA, elGamal, and elliptic curve. 
Because no parameter is required in a parameter-free algorithm, anyone who has sRes and sInput can easily utilize the identical algorithm Fun to verify whether the data sInput is a tamper. In contrast, because the parameter is owned by a specific person, the data sRes only can be generated by this specific person and verified by identical or different persons. Especially, in an asymmetric algorithm, it will be very difficult for a specific person who has the private key to generate the data sRes to deny that he or she signed the data. Thus, the output sRes encrypted with a private key of an asymmetric algorithm is usually called a signature. 
Similar Prior Art of Blockchain
As the name indicates, a blockchain is constructed with multiple components, named "blocks," and these multiple components are connected to each other in the form of a "chain." The structure of multiple components connected in a chain form is not a unique or newly created concept. For example, a passbook can be considered as such a type of structure, in which each row can be considered as one block and all of the rows (blocks) are connected and arranged chronologically. 
This type of passbook is relatively secure. For example, a forged passbook with several continuous or non-continuous modified rows would be easily found, because in order to forge such a document one would need not only to carefully manage consistency of the data in the columns "deposit," "withdrawal," and "balance," but also carefully adjust the difference of the font and/or the type of ink used between the original rows and modified rows. 
Another example of such multiple components connected in a chain can be the mechanism of a "chain of trust," which is utilized in communication networks, more specific, in G.509 [1]. Fig. 1 shows an illustrated scheme of a chain of trust utilized to create multiple certificates, wherein certificates may be considered as blocks in a blockchain. A certificate, except for the first certificate (root certificate), has at least four fields, i.e., owner’s name, owner’s public key, issuer’s name, and issuer’s signature. The fields of the owner’s name and owner’s public key are utilized to present that this certificate is created for the specific user and the specific user’s public key. The field of the issuer’s name indicates who issued the certificate. After the issuer verifies the identity of the owner with another method, the issuer will use its (the issuer’s) own private key to generate a signature for the owner’s public key. Then the generated signature is filled into the field of the issuer’s signature. During this process, the public key of the issuer at a lower level may be verified by another issuer at a higher level, for example the root issuer, with the same process. Hence, the field of the owner’s name of the higher issuer’s certificate is identical to the issuer’s name field of the lower level issuer. It should be noted that because the root certificate is not issued from another certificate, the owner’s and the issuer’s name fields are merged into a single name field. Besides, a user may not prepare a public key and its corresponding private key because the issuer may generate such public and private pair of keys. 
In the above example, one reference chain is constructed from the current issuer’s name to the previous owner’s name. Hence this reference chain can be utilized to trace who issued certificates. However, one potential weakness is that the fields of the owner’s name and issuer’s name are presented as a plaintext without being protected by any cryptology algorithm. Hence, anyone can modify the content of the issuer’s name and forge a certificate. Besides, the correct operation of the chain of trust scheme is based on the assumption that all of the issuers are reliable. However, this assumption may not necessarily be true, because several fake certificates may be issued after the computer of the issuer is hacked. 
Overview of Bitcoin
In summary, bitcoin has the following interesting technical features: PrevBlockHash, Nonce, multiple loggers, transparence, and renewed scheme. 
Fig. 2 shows the scheme diagram of bitcoin and/or a blockchain [2-5]. To easily compare Figs 1 and 2, a payload section is defined. Comparing Fig. 1 and Fig. 2, the data in the payload section in Fig. 2 is TX, the denotation of a transaction which records how the currency is transacted during a specific time period. The specific time period is relative to the timestamp in the field "Time." Besides, Fig. 2 also shows a similar reference chain which is constructed from the "PrevBlockHash" field in the current block to the "BlockHash" field in the previous block. 
One of the important features in Fig. 2 is that all of the data, except for the field BlockHash, is protected by a cryptology algorithm, such as a hash function. More specific, the data in the payload section may not directly combine the data in the header section, and their combined data is processed by a cryptology algorithm, because the computation and storage cost may be too expansive. Thus, bitcoin proposed a Merkle tree structure, as shown in Fig. 3, wherein transaction hash values are paired and hashed to provide the value in the MerkleRoot field in the header section. Because any modification applied to the payload section will simultaneously modify the hash result, which is duplicated in the field of BlockHash in the current block and in the field of PrevBlockHash in the next block. As a result, if a hacker plans to modify a previous transaction, for example a transaction from two months ago, the hacker would need to modify all blocks, or at least their fields of hashing, starting from the block with the target transaction up to the current block. 
As shown in Fig. 2, the fields in the header section include BlockHash, PrevBlockHash, Nonce, Time, and MerkleRoot. Another important feature of bitcoin is nonce, a random number. A block is considered to be good if the hash value of the block is smaller than a target value. Thus, large numbers of nonces have to be tested repeatedly and it has to be checked whether the corresponding generated hash value is smaller than the target value. Hence, this process will slow down generation of a fake block. 
In Fig. 1, the data structures are stored in central servers. Thus, if any one of the servers is hacked, the data may be forged. Besides, for security reasons, only certain staff has the right to access the servers or review/modify the data in the servers. However, in case of a delayed reaction, a hacker may only be detected after several days or months, which constitutes a severe security issue.
As a response, bitcoin proposes a peer-to-peer architecture to replace the central architecture. Every node computer will own copied data, resulting in a large number of identical copies. Thus, it is difficult for a hacker to modify such large numbers of data within a limited time interval. Besides, another advantage of such peer-to peer architecture is transparence because every node computer can accept the data. 
In bitcoin, several transactions in a time interval are collected as a payload in block. These transactions are not limited to one specific person but may come from several different persons. To ensure one person has enough assets before transfer, as well as protect one person form transferring an identical asset to different accounts ("double-spending problem"), these transactions need to be verified before they are combined as the final block, i.e., the current block. 
To prevent the witness or the verification being a hacker, bitcoin applies a verification policy named "stack as work," which in fact is a game in which participants compete to find a nonce to the corresponding hash value smaller than a target value. Thus, the node computers may also be devoted to verifying all the transactions in one block. After a nonce candidate is claimed, the other node computer will verify whether the nonce candidate can make the hash result of the current block match the required condition. Only if the majority of the node computers agree that the nonce is suitable, this processing block is deemed a real one and attached to the tile of the chain structure. As long as a hacker cannot control the majority of the computer resources, bitcoin is safe. In other words, the devotion of the node computers is a key element contributing to the success of bitcoin. 
To encourage participation, bitcoin proposed a feedback reward mechanism. The first one who finds a real nonce candidate receives not only transaction fees but also reward coins. Of course, to maintain the security and to prevent currency devaluation, the target value will be adjusted every 2,016 blocks, and the amount of reward coins will be reduced by half every 210,000 blocks. The target value is adjusted to meet a verification rate of approximately one block every 10 minutes [2]. 
In summary, in the author’s opinion, the success of bitcoin can be attributed to three main reasons: chain structure to protect the payload from being forged, the nonce to ensure the computing cost is high, and the reward system to encourage a large number of participants. 
Overview of Small Contracts
In bitcoin, a transaction is initialized by a real human being. However, technologies in the field of "High Frequency Trading" have been applied to the stock mark. The key feature of HFT is that the transaction is initialized and completed by a computer program instead by a human being, although the human being may predetermine trigger situations. Thus, the skills used in bitcoin are modified such that a program code is included in the blockchain. When certain conditions are met, the program code will be executed automatically in a prescribed manner on every node in the network [11, 12, 14]. 
Non-financial Applications of Blockchain
Bitcoin and blockchain originated in the field of financial technology or “FinTech,” so naturally most of their associated applications focus on the financial field. It is believed that more than 600 crypto currencies have been created, the most famous ones being Ethereum, Monero, and Ripple [4]. However, some researchers tried to extend the applications of blockchain to other fields. 
For example, Nir Kshetri introduced several examples in which blockchains are applied to e-voting [6]. Besides, Muhamed Turkanović et al. applied blockchain to higher education credits [7], Yinghui Zhang et al. applied blockchain to keyword search of encrypted data [8], Nicolas Herbaut et al. applied blockchain to video delivery [10], Weizhi Meng et al. applied blockchain to intrusion detection, Fabrizio Lamberti et al. applied blockchain to car insurance [14], and Jie Zhang et al. applied blockchain to healthcare [15]. It should be noted that the implementations of the latter two examples relate to a portable device, an embedded system, and/or an integrated circuit. Thus they can be considered to belong to the field of Internet of Things (IoT). In addition, several researchers indicate that blockchains can be applied to the field of IoT [11, 12, 13]. Thus, the connection between blockchains and IoT is an important trend. For more applications of blockchain and the combination between blockchain and IoT please refer to [4, 11]. 
In addition, the improvement of hardware architectures is an important research topic. For more details please refer to [16, 17]. 
Challenges of Non-Financial Applications
Among the above non-financial applications, some applications are concerned with how the data presented in blockchain are explained, and they do not fundamentally modify the technical scheme. For example, in the application of higher education credits, the ECTX tokens represent the equivalent to a student’s credits for completed courses, and each student will own a blockchain wallet to collect CTX tokens. Every time a student completes a course, his/her home higher education institution will transfer an appropriate number of ECTX tokens to his/her blockchain address [7]. Similarly, in “e-voting applications,” a manager issues each voter a “wallet” containing a user credential. Each voter receives a single “coin” representing one opportunity to vote. Casting a vote transfers the voter’s coin to a candidate’s wallet [6]. 
The successes of blockchain can be attributed to the high computing resource requirements of proof-of-work, but this feature may prevent blockchains from being combined with IoT devices. This is because IoT devices have limited computation and storing abilities [13], and this limitation may result from safety policies [15]. Besides, one node may store a local copy of 105 GBytes for bitcoin [4]. 
To resolve the computing issue, some researchers arranged for IoT devices to only have the ability to transmit its data to a node of a blockchain network. For example, Jie Zhang et al. propose that the sensor devices can only use a first communication protocol to communicate with a coordinator device, for example a mobile phone, and then the coordinator device can broadcast the data to miner nodes to add the data into a blockchain [15]. Thus, the computing and storing resources for sensor devices are reduced. Similarly, Oscar Novo et al. proposed that IoT devices transmit data to management hub devices, and the management hub devices forward the data to the blockchain network [13]. 
Regarding the storing space issue, although other blockchain architectures have been proposed, it is believed that these solutions are too difficult to be applied to most IoT devices [13]. Besides, the fees for utilizing a blockchain may depend on the data volume transmitted to the blockchain. Thus, the original raw data may not be stored in the blockchain. In the healthcare application proposed by Jie Zhang et al., the data stored in the blockchain are addresses rather than health data [15]. Similar, Fabrizio Lamberti et al. proposed to transmit the hash values of vehicle’s pictures, instead of the raw pictures, in case of theft, fire, and weather events [14]. In addition, Fabrizio Lamberti et al. utilized a weaker hash function MD5 to generate the photo's hash value in order to reduce the bit number of a hash value [14]. The data may not automatically be periodically transmitted, but the data may be transmitted when a user manually decides to transmit it. Alternatively, the data may also be transmitted automatically when some specific situations occur, for example, when the vehicle leaves or enters an area specified by the user [14]. 
In addition, some researchers further indicated that blockchain may take some un-skippable time to propagate the data into the other nodes of a blockchain network [2]. For example, bitcoin transactions can take up to 10 min and 12 sec in Ethereum. Thus, real-time applications are not suitable. Regarding this issue, Oscar Novo et al. proposed an expiration date parameter and/or to higher transaction fees [13]. Fabrizio Lamberti et al. also proposed to pay higher transaction fees, and proposed to consider building a private blockchain [14]. 
Challenges for Patent Attorneys
Generally speaking, a blockchain is a peer-to-peer network architecture. An obvious challenge for patent attorneys is how to propose suitable protection claims for the interactions between the different entities or nodes, because those nodes may be manufactured, sold, or utilized by different companies or entities. For example, in the application of video delivery proposed by Nicolas Herbaut et al., three types of users (end user, content provider, and technical enabler), three types of blockchains (delivery monitoring blockchain, content provision blockchain, content brokering blockchain), and three types of smart contracts (content brokering contract, content licensing contract, content delivering contract) are involved [10]. In the IoT application proposed by Oscar Novo, there are five types of devices (IoT device, management hub, miner, agent node, manager) [13]. These different components are necessary to be defined because of their different behaviors and/or characters. Thus, it is a great challenge for patent attorneys to identify the critical components so as to define the scope of claims. 
Current Patent Situation in Taiwan
Research is also conducted based on data available from the Taiwan IPO database by searching the key words bitcoin and blockchain and their Chinese translations in titles, specifications, and claims. Because bitcoin was released on January 03, 2009, only data after said date is analyzed. Until September 14, 2018, there are seventy-eight early published invention patent applications, and thirty-seven invention and utility model patents were gazetted. In the early-published recording data, applicants with a comparatively large numbers of early-published invention applications are NCHAIN HOLDINGS LIMITED (AG, 10), MODERNITY FINANCIAL HOLDINGS, LTD. (KY, 9), NCHAIN HOLDINGS LIMITED (AG, 7), APPLE INC. (US, 7), 21, INC. (US, 5), FUBON FINANCIAL HOLDING CO. (TW, 5), wherein the numbers of FUBON FINANCIAL HOLDING CO. include the applications of FUBON INSURANCE CO., LTD. because they are related companies. In the gazetted recording data, applicants with a comparatively large numbers of gazetted invention and utility model patents are MODERNITY FINANCIAL HOLDINGS, LTD. (KY, 5), FUBON FINANCIAL HOLDING CO. (TW, 5), and APPLE INC. (US, 2). It should be noted that the first code in parentheses is the nation code of the first applicants, and the second number is the early published or gazetted case numbers. 
A more detailed analysis was conducted to determine the ratio of applicants being banks, insurance companies, and/or financial companies. In the early-published recording data, twenty-three percent (18/78) of cases belonged to the aforementioned types of companies. In the gazetted recording data, forty-nine percent (18/37) of cases belonged to the above categories. As a result, not all of the applicants are banks, insurance companies, and/or financial companies, so there appears to be additional focus on non-financial applications. On the other hand, it is observed that APPLE INC. is organizing patents in the field of blockchain. Besides, the next steps of INTEL CORPORATION (US) and SAMSUNG ELECTRONICS CO., LTD. (KR) should be observed although they respectively have only two and one early-published invention applications listed in the Taiwan IPO database. 
Conclusion
Bitcoin/blockchain proposes mechanisms to resolve several drawbacks, but such mechanisms may result in new limitations to be resolved when applications of bitcoin/blockchain are extended into different fields, and these solutions should be patentable. 
References
[1] "Root certificate," Wikipedia, website: https://en.wikipedia.org/wiki/Root_certificate, retrieved on Sep. 19, 2018.
[2] Florian Tschorsch and Björn Scheuermann, "Bitcoin and Beyond: A Technical Survey on Decentralized Digital Currencies," IEEE COMMUNICATIONS SURVEYS & TUTORIALS, vol. 18, no. 3, pp. 2084-2123, 2016.
[3] Ittay Eyal, "Blockchain Technology: Transforming Libertarian Cryptocurrency Dreams to Finance and Banking Realities," Computer, vol. 50, no. 9, pp. 38-49, 2017.
[4] Valentina Gatteschi, Fabrizio Lamberti, Claudio Demartini, Chiara Pranteda and Víctor Santamaría, "To Blockchain or Not to Blockchain: That Is the Question," IT Professional, vol. 20, no. 2, pp. 62-74, Mar./Apr. 2018.
[5] Tomaso Aste and Paolo Tasca, "Blockchain Technologies: The Foreseeable Impact on Society and Industry," Computer, vol. 50, no.9, pp. 18-28, 2017.
[6] Nir Kshetri and Jeffrey Voas, "Blockchain-Enabled E-Voting," IEEE Software , vol. 35, no. 5, pp. 95-99, 2018.
[7] Muhamed Turkanović , Marko Hölbl, Kristjan Košič , Marjan Heričko, and Aida Kamišalić, "EduCTX: A Blockchain-Based Higher Education Credit Platform," IEEE Access, vol. 6, pp. 5112-5127, 2018.
[8] Yinghui Zhang, Robert H. Deng, Jiangang Shu, Kan Yang, and Dong Zheng, "TKSE: Trustworthy Keyword Search Over Encrypted Data With Two-Side Verifiability via Blockchain," IEEE Access, vol. 6, pp. 31077-31087, 2018.
[9] Weizhi Meng, Elmar Wolfgang Tischhauser, Qingju Wang, Yu Wang, and Jinguang Han, "When Intrusion Detection Meets Blockchain Technology: A Review," IEEE Access, vol.6, pp. 10179-10188, 2018.
[10] Nicolas Herbaut and Daniel Negru, "A Model for Collaborative Blockchain-Based Video Delivery Relying on Advanced Network Services Chains," IEEE Communications Magazine, pp. 70-76, September 2017.
[11] Tiago M. Fernández-Caramés and Paula Fraga-Lamas, "A Review on the Use of Blockchain for the Internet of Things," IEEE Access, vol.6, pp. 32979-33001, 2018.
[12] Konstantinos Christidis and Michael Devetsikiotis, "Blockchains and Smart Contracts for the Internet of Things." IEEE Access, vol. 4, pp. 2292-2303, 2016.
[13], Oscar Novo, "Blockchain Meets IoT: An Architecture for Scalable Access Management in IoT," IEEE INTERNET OF THINGS JOURNAL, vol. 5, no. 2, PP. 1184-1195, APRIL 2018.
[14] Fabrizio Lamberti, Valentina Gatteschi, Claudio Demartini, Matteo Pelissier, Alfonso Gómez, and Victor Santamaria, "Blockchains Can Work for Car Insurance," IEEE Consumer Electronics Magazine, vol. 7, no. 4, pp. 72-81, July 2018.
[15] Jie Zhang, Nian Xue, and Xin Huang, "A Secure System For Pervasive Social Network-Based Healthcare," IEEE Access, vol. 4, pp. 9239-9250, 2016.
[16] "GOLDSTRIKE 1: COINTERRA’S FIRST-GENERATION CRYPTOCURRENCY MINING PROCESSOR FOR BITCOIN," IEEE Micro, vol. 35, no. 2, pp. 68-76, Mar.-Apr. 2015.
[17] Michael Bedford Taylor, "The Evolution of Bitcoin Hardware," Taylor Computer, vol. 55, no. 9, pp. 58-66, 2017.