（原文譯文都有）用傳輸數據記錄器評估區(qū)塊鏈技術

1、<p>　　出處在后面譯文部分?。?！</p><p>　　Assessing block chain technology for Transport Data Logger</p><p>　　Björn Johansson</p><p>　　List of acronyms</p><p>　　APIApplic

2、ation Programming Interface</p><p>　　BFTByzantine Fault Tolerance (alt. Byzantine Fault Tolerant) CACertificate Authority</p><p>　　CPUCentral Processing Unit</p><p>　　HTTPSHyper

3、text Transfer Protocol Secure MSPMembership Service Provider PBFTPractical Byzantine Fault Tolerance PKIPublic Key Infrastructure</p><p>　　PoETProof of Elapsed Time</p><p>　　RESTRepresenta

4、tional State Transfer SDKSoftware Development Kit</p><p>　　SGXSoftware Guard Extensions</p><p>　　SHASecure Hash Algorithm</p><p>　　SSHSecure Shell</p><p>　　TDLTran

5、sport Data Logger</p><p>　　TLSTransport Layer Security</p><p>　　Popular Science Summary:</p><p>　　Storing and attributing log data using blockchain technology</p><p>　

6、　Blockchain technology is incredibly young, even the most mature of the blockchain technologies are either functionally complete but very simple or lack important features. The thesis has assessed blockchain technology f

7、or implementation with the Bosch Transport Data Logger, TDL.</p><p>　　The TDL is a sensor-equipped logging device with a companion smartphone app. The device is fastened to packages during transportation and

8、 logs how the packages have been handled while in the care of post and transport companies. For example, unacceptable temperatures, humidity and manhandling are logged.</p><p>　　The two main types of blockch

9、ain technology, permissionless and permissioned, are found to have inherent trade-offs between open/closed participation and poor/good performance. Hyperledger Fabric, most mature of the permissioned blockchain systems,

10、was used to build a TDL Blockchain System proof of concept. Log data and other information from and about the TDL devices, transferred by the transport company employees using the companion app, are stored on the blockch

11、ain.</p><p>　　Generally, just saving sensor values is not enough to accomplish anything. To give credence to the logged data and to be able confidently to act upon it, the data must be made trustworthy. In a

12、n environment where data is generated while the sensor device is in the hands of different organizations, trustworthy connections between each data point and the organization responsible must be made. This to prevent bla

13、me-shifting of the type “it was like that when I got it”. To bring historical and docum</p><p>　　The thesis can be used by Bosch to learn about blockchain technology and its applicability for the TDL. If the

14、y or others want to extrapolate they could use the thesis results to draw conclusions for other similar devices and systems. The general thesis recommendation is two-fold. If an organization simply wants to implement blo

15、ckchain technology applications and not help develop the underlying technology the blockchain technology sector is not yet ready for them, otherwise now is a good time to</p><p>　　Introduction</p><

16、;p>　　As the steps in a transport chain increases the more distrust between transport chain actors (companies) can be expected as those who commit errors while handling sensitive goods would most likely want to shift t

17、he blame to another actor in the transport chain. The Bosch Transport Data Logger, TDL, was created to determine what conditions a package equipped with a TDL device is exposed to during transport.</p><p>　　

18、This thesis will look at how blockchain technology can be used to make the data generated by the TDL more reliable and how it can be used to pinpoint what transport chain actor had responsibility for the package when any

19、 sort of unacceptable conditions were logged.</p><p>　　1.1 Background</p><p>　　The Transport Data Logger system consists of a sensor-equipped device that interacts with a companion smartphone

20、app. The device is fastened to goods during transport and set up using the app to tolerate certain threshold sensor values including temperature, impacts and humidity (and more), with readings exceeding those values bein

21、g logged as violations. The recipient can then connect to the TDL device with the smartphone app upon delivery of the goods and then view (and share) the log data.</p><p>　　The problem of “it was like this w

22、hen I got it” can potentially be solved by logging if and when conditions damaging to the goods occur at any time during transportation, recording every handoff between actors in the transport chain and protecting the re

23、corded data from manipulation.</p><p>　　In its current form the TDL system lacks cryptographic protection or attribution of violation data to specific transport chain actors, other than in the form of manual

24、ly comparing timestamps.</p><p>　　Data verification today relies heavily on having a common trusted third party that vouches for the dependability, source and integrity of that data.</p><p>　　Bl

25、ockchain technology promises to provide network-distributed, decentralized and immutable data storage and transaction conduction. This would eliminate the need to completely trust a single third party to verify the integ

26、rity or existence of some data. Instead trust is put in the collective that is made up of entities collaborating in a blockchain network.</p><p>　　Some form of consensus building methodology is needed to mak

27、e sure that the blockchains at all entities nodes are identical, making the figurative collective blockchain stable. All good consensus building methods are tamper resistant in that for tampering to occur, many different

28、 actors on the blockchain network must collaborate. With a traditional single centralized database only direct changes to the database would be needed for tampering to succeed.</p><p>　　The brute-force natur

29、e of creating new blocks for proof-of-work consensus blockchains has led to that form of block creation being called “mining”.</p><p>　　1.2.Problem Statements</p><p>　　These three questions ar

30、e good to keep in the reader’s mind as they are the guiding lights of the thesis report.</p><p>　　Question 1. What are the different strengths and weaknesses of permissioned and permissionless blockchains?&l

31、t;/p><p>　　To decide what type of blockchain system should be used for the TDL Blockchain System, the defining characteristics of the two main types of blockchain technology must be determined. Identifying how

32、blockchain technology works is also important in being able to answer the other questions in the problem statement.</p><p>　　Question 2. How can blockchain technology be used to increase the reliability of t

33、he Transport Data Logger data?</p><p>　　Question 3. How can blockchain technology be used to attribute specific data points from the Transport Data Logger to specific transport chain actors?</p><p

34、>　　To be able to use the data generated by the TDL to make well-informed decisions that data must be determined to be reliable, meaning accurate and unmanipulated. It must also be possible to directly attribute specif

35、ic data to transport chain actors, preventing them from shifting blame away from themselves.</p><p>　　1.3.Method</p><p>　　This thesis is exploratory in nature, at the beginning a vague approxi

36、mation of what would later become Question 2 and Question 3 as well as the intent to develop a proof of concept was used to create a schedule and method outline (a project plan).</p><p>　　As the author could

37、 not for several very important reasons commit any more time to the thesis than the standard one semester this thesis was from the start conducted to fit inside of those 20 weeks (5 months). Other thesis projects known t

38、o the author had exhibited a tendency to expand time to fit the scope and to prevent that from happening care was taken from the start to create and stick to a reasonable schedule, method and scope. The schedule outline

39、looked roughly like the following: develo</p><p>　　Fig. 1. The method outline showing the 5 months of the project and the phases they contained. This is a simplified version of reality as for example outline

40、 drafts of the thesis were created during the research phase and research in one form or another was conducted during almost all project phases.</p><p>　　The method outline was to spend the first month explo

41、ring the blockchain technology field and doing much of the necessary research for the thesis. This included Lund University Libraries1 and Swedish thesis report2 searches for terms like, or related to, “blockchain”. It a

42、lso included looking at interesting references and technologies from the results of those searches. The initial research phase also included looking through Bosch internal development documentation for the TDL, exploring

43、 the TD</p><p>　　Once the main use case for the TDL Blockchain System was supported the development phase was over. It was followed by a month-long thesis report writing phase. Writing the thesis was mostly

44、done chapter by chapter. An outline of the thesis chapters and their contents had been worked on during the research phase and writing continued from that as well as the lessons learned during development. While focus wa

45、s kept on one chapter at a time the outline of the entire thesis was still worked on all t</p><p>　　The final month of the project was spent finishing and polishing the thesis report based on feedback, prepa

46、ring and performing presentations and working to complete all other requirements associated with the final phase of a thesis project. Opposition could unfortunately not be arranged before the end of the five months and a

47、s such spilled into the subsequent semester. This was unfortunate but ultimately acceptable.</p><p>　　All throughout the thesis project a daily diary was kept and weekly status reports to the supervisors wer

48、e made. This, along with the rigorous use and iterative development of the project plan containing the schedule and decisions around method and scope, was crucial in keeping the project on track and on time.</p>&

49、lt;p>　　1.4. Related Work</p><p>　　Jeppsson and Olsson did a Master’s thesis on the usage of blockchains for tracking goods during transportation [1]. Their thesis focuses on the impact to the transport

50、chain actors, offering an excellent complement to this thesis which focuses more on the security and technical aspects of such a blockchain system. Their thesis is highly recommended reading.</p><p>　　In ano

51、ther related Master’s thesis Jansson and Petersen developed a framework for evaluating blockchain as a supply chain traceability system [2]. While their framework is not used in this thesis important topics brought up by

52、 them are addressed such as the difference between documental and historical accuracy.</p><p>　　Yli-Huumo et al mapped out existing areas of blockchain research up until 2015 [3]. They concluded that 80% of

53、research focused on Bitcoin specifically instead of the general field of blockchain technology. They also</p><p>　　summarize many areas of concern for blockchain technology such as throughput, latency, reso

54、urce use, attack vectors and privacy.</p><p>　　Important aspects of performance and scaling in blockchain systems is addressed by Scherer [4]. Many of his results, as supported by other sources, are importan

55、t for highlighting the differences between permissioned and permissionless blockchain technology in this thesis.</p><p>　　There is quite a bit of research and development going into blockchain applicability

56、for supply chains as the area is considered “promising” [5]. Most of that work is primarily focused on moving the physical paperwork surrounding transportation to a digital system based on blockchain technology, and to t

57、hen be able to trace a product through the entire supply chain with that blockchain data. As examples, [1] and [2] are already mentioned. See also [5, p. 23], [5] itself, and from industry see fo</p><p>　　[6

58、] and [7]. It is notable that much of this work is less than a year old, speaking to the emergent nature of blockchain technology.</p><p>　　1.5. Thesis Outline</p><p>　　This thesis is divided

59、into three main parts. 0 deals with cryptography, blockchain technology and explains the current TDL security solution. Part II deals with the TDL Blockchain System and its proof-of-concept implementation. Part III conta

60、ins a discussion of blockchain technology and the TDL Blockchain System as well as potential future work and the thesis summary.</p><p>　　Cryptography</p><p>　　An understanding of the basic conc

61、epts of modern cryptography is required understanding for many important aspects of blockchain technology. The field of modern cryptography is highly mathematics-based but those specifics are not required understanding f

62、or blockchain technology.</p><p>　　To see roughly what section of cryptography aids understanding of what section of other cryptography and blockchain technology see Fig. 2.</p><p>　　Fig. 2.The

63、 cryptography areas on the left help in understand the areas of cryptography and blockchain technology on the right.</p><p>　　Because of the basic nature of this section on cryptography the information is la

64、rgely sourced from the lecture notes for EIT060 Computer Security held in 2017 at Lund University, Faculty of Engineering [8].</p><p>　　In cryptographic scenarios, it is common to use “Alice” and “Bob” to de

65、note trusted parties wishing to communicate in some way, and “Eve” as someone wishing to eavesdrop or otherwise compromise the scenario.</p><p>　　2.1.Kerckhoffs’ principle</p><p>　　Cryptographi

66、c algorithms use keys to protect data, and Kerckhoffs’ principle states that a cryptosystem should remain secure even if everything about it, excluding its key, is public knowledge [9].</p><p>　　Related to K

67、erckhoffs’ principle is the general principle of only using cryptographic primitives that have been thoroughly tested using extensive research, also known as “don’t roll your own crypto” [10] [11].</p><p>　　

68、2.2.Hashing</p><p>　　Hash functions, hash(𝑚) = ?, have two defining properties. Firstly, they should map a bit message of arbitrary length 𝑚 to a fixed-length output ?, see Fig. 3. Secondly,

69、 they should be computationally light.</p><p>　　Fig. 3. Messages of varying length are run through the cryptographic hash function SHA3-256 generating hash values of the fixed length 256 bits.</p><

70、;p>　　To make hash functions suitable for use in cryptography, three additional properties are needed.</p><p>　　Pre-image resistance: given a hash value ?, it is infeasible to find</p><p>　　&#

71、119898; such that hash(??) = ?.</p><p>　　Second pre-image resistance: given a message 𝑚, it is infeasible to find ??’ such that hash(𝑚) = hash(??’).</p><p>　　Collision resistance

72、: it is infeasible to find 𝑚 and ??’ such that</p><p>　　hash(𝑚) = hash(??’).</p><p>　　Cryptographic hash functions provide small representations of potentially much larger data.

73、This is very valuable when signing data as without representing the data with its hash value the signature would have to be just as large as the data itself. Section 2.5 deals with signing data using asymmetric cryptogra

74、phy.</p><p>　　2.3.Symmetric cryptography</p><p>　　Cryptography has historically only consisted of what we today call symmetric cryptography. It is used to send encrypted messages back and forth

75、 between trusted parties sharing a common secret key. This key is, as its name suggests, kept secret from everyone else.</p><p>　　Alice and Bob trust each other and share a secret key 𝑘. Alice encryp

76、ts a plaintext message 𝑚 using the symmetric encryption function encrypt(𝑚, 𝑘) and sends the resulting ciphertext 𝑐 to Bob over some communication channel. After receiving the ciphertext B

77、ob uses the corresponding decryption function decrypt(𝑐, 𝑘) providing it the same secret key that Alice encrypted the message with. The result, the original message 𝑚, can now be read by Bob. This

78、 is illustrated in Fig. 4.</p><p>　　Fig. 4. Alice encrypts the message “HELLO FROM ALICE” using a secret key and sends it to Bob who uses that same key to decrypt the ciphertext “NVZHF JRJQ RDVQA” into the o

79、riginal message. Eve, who only has access to the ciphertext and not the key, is unable to access the message.</p><p>　　The main advantage of symmetric encryption is speed and resource use. It is, in general,

80、 computationally lighter to encrypt and decrypt data with symmetric encryption than it is to use asymmetric cryptography, described in Section 2.4.</p><p>　　2.3.1.Exchanging secret keys</p><p>

81、　　A problem is the need to share the secret key outside of the cryptosystem, which creates an attack vector. When describing a symmetric encryption system, it is generally assumed that the key is already shared between t

82、he trusted parties, leading to the alternate name “pre-shared key”. When describing symmetric encryption schemes, keys are assumed to be distributed in a safe manner outside the cryptosystem.</p><p>　　It is

83、common to use the slower asymmetric cryptography to exchange a secret key and then switch to symmetric cryptography for the rest of the session.</p><p>　　2.4.Asymmetric cryptography</p><p>　　In

84、stead of being based on sharing a common secret key, asymmetric cryptography is based on the concept of having a pair of mathematically linked keys. Every participant holds one of these key pairs, keeping their “private

85、key” secret while actively sharing their “public key”.</p><p>　　One of the most famous asymmetric encryption schemes is RSA, named after its three creators Rivest, Shamir, and Adleman. RSA uses a form of tra

86、pdoor function, taking advantage of the computational difficulty of finding the factors of large numbers. It is easy to multiply two large prime numbers to generate a very large number, but to find the two prime factors

87、only knowing the large number is very difficult.</p><p>　　Asymmetric cryptography can, in addition to its use in encryption, also be used to create digital signatures that ties data to a specific source. Sig

88、natures can also be used to validate that the data has not been changed since it was signed.</p><p>　　2.4.1.Key pair</p><p>　　The public and private keys are cryptographically linked in such a

89、way that if you encrypt a message with the public key only someone who holds the private key can decrypt it. It is also important that it is infeasible to find the private key if you’re given the public key. These proper

90、ties make it easy to verify that someone has access to a certain private key, you simply need to encrypt some random data with a public key and if they can decrypt it then you know that they have the associated pr</p&

91、gt;<p>　　Note that the names “public” and “private” refer to how the keys are used.</p><p>　　2.4.2.Asymmetric encryption and decryption</p><p>　　A message encrypted with the public key c

92、an only be decrypted with the private key.</p><p>　　Alice has generated her key pair consisting of her public key and private key. Bob has, in turn, generated his own key pair . Alice and Bob have both share

93、d their public keys and with one another in such a way that they both know that the respective key belongs to the other.</p><p>　　Alice wants to send a plaintext message to Bob so she uses Bob’s public key

94、 to encrypt the message into ciphertext using . She then sends to Bob who decrypts the message using. This is illustrated in Fig. 5.</p><p>　　Fig. 5. Alice encrypts a message “HELLO FROM ALICE” using Bob’s

95、 public key. She sends the resulting ciphertext to Bob who decrypts it using his private key. Eve who knows both the ciphertext and Bob’s public key is unable to decrypt the message as she does not know Bob’s private key

96、.</p><p>　　It should be noted that the message is in practice not encrypted/decrypted immediately, some sort of pre- and post-processing should be implemented to prevent the identical messages from being enc

97、rypted to the same ciphertext, a fact that can otherwise be used by Eve to gain information about the messages sent [12].</p><p>　　2.5.Signing</p><p>　　To prove the origin of a message it is co

98、mmon to send a digital signature along with the message itself.</p><p>　　The signing process begins with Alice in possession of her own key pair, and her having shared her public key in such a way that every

99、one knows that she is really the owner of it. Alice wants to send a message 𝑚 to Bob so that Bob can verify that it came from her.</p><p>　　Alice produces a hash value ? = hash(𝑚) of the mess