14-BlockChain Puzzles and Concerns

1 .CS5412/Lecture 14 Blockchains for I o T (Part 2) Ken Birman CS5412 Spring 2019 http://www.cs.cornell.edu/courses/cs5412/2019sp 1

2 .Let’s look at a Blockchain created specifically for IoT Cornell has a new “smart farms” research effort. Led by Susan McCouch , Hakim Weatherspoon and Abe Strouck . One early accomplishment: Vegvisir , a BlockChain specifically for agriculture. http://www.cs.cornell.edu/courses/cs5412/2019sp 2

3 .Vegvisir Slides from Robbert van Renesse Talk presented at ICDCS 2018 http://www.cs.cornell.edu/courses/cs5412/2019sp 3

4 .A Blockchain for the Food Supply Chain Robbert van Renesse joint work with Hakim Weatherspoon, Danny Adams, Kolbeinn Karlsson, and Stephen B. Wicker Initiative for Crypto-Currencies and Contracts (IC3) Cornell Digital Agriculture Initiative

5 .Blockchain’s Promise Promises Global currency Smart contracts Notarization Accountability …

6 .A Replicated Ledger of Transactions give 10 bitcoins to Jane Ledger Judy (owns 15 bitcoin ) Joe (owns 1 bitcoin ) give 10 bitcoins to Joe give 3 bitcoins to Judy Jane

7 .Smart Contracts Smart contracts are executable programs on the BlockChain , take input from the BlockChain , and produce output on the BlockChain Main use: automated escrow , where disbursement depends on agreed upon conditions Caution: Smart Contracts have been found to be prone to (very expensive) bugs 7

8 .Potential Use Cases Killer app: cryptocurrencies Other potential uses: Reduce opaqueness of supply chains One “trustless” place for all transactions along the way Improvements over paper-based systems and many disjoint databases Eliminate middlemen Why does farmer make so little and consumer pay so much? Reduce fraud India, Russia, Sweden, Georgia … are building blockchain -based land registries to fight “land fraud” and simplify international property transactions 8

9 .For the Food Supply Chain? Supply chain management Walmart is building one for the food supply chain Food safety: fast identification of tainted foods Consumers are demanding more information about the products they buy (organic, fair trade, … ) Simplify international transactions Help farmers Want to know what happens to their products for fair pricing What products should they be producing? Reduce food scandals illegal production, misrepresentation, loss and waste, … 9

10 .Industrial uptake? ripe.io : A company that is building a “ blockchain of food” with IoT interfaces Walmart : partnered with IBM and Tsinghua to identify sources of contaminated products and speed up recall But today’s blockchain technology may not be appropriate for all use cases too dependent on availability of plentiful power, networking, and storage 10

11 .Desired Blockchain properties Performance : High Throughput, Low Latency Energy-Efficient Security : Always available for reading (verifying) and appending Fair Tamperproof (Integrity) Possibly confidentiality as well No Single Administrative Domain no need to trust a single provider Open membership (or not)

12 .Open Membership is Hard Traditional secure logs are based on voting Members vote on which transactions to add to the log and in what order Problem: “Sybil” or impersonation attacks a participant may try to vote multiple times with closed membership, cryptographic signatures can identify the source of a vote with open membership, anybody can create identities and that way vote many times 12

13 .Permissionless vs Permissioned Blockchains 13 Permissionless Permissioned Approach Competitive Cooperative Basic technique Proof-of-Resource Voting Membership Open Closed Energy-efficiency Often terrible Excellent Transaction rate At best hundreds / sec Many thousands per second Transaction latency As high as many minutes Less than a second

14 .Bitcoin Blockchain Permissionless , open membership Proof-of-Work There are thousands of Bitcoin miners they use ASIC hardware to compute SHA256 hashes use about more energy than the country of Denmark Overall rate is a few transactions per second 14

15 .The Blockchain Ledger   HASH( ) < target nonce “ cryptopuzzle ”

16 .The Blockchain

17 .The Blockchain

18 .The Blockchain Exponentially distributed rate of new blocks, with constant mean interval target automatically adjusted every 2016 blocks so that mean interval is 10 minutes

19 .Incentives for Mining Prize: “Minting” Transaction Fees Wins proportional to computation power

20 .Forks Two blocks “mined” at approximately the same time by two different miners

21 .Fork Resolution Longest chain wins Transactions on short chain are reverted

22 .Fork Resolution A transaction is confirmed when it is buried “deep enough” (typically 6 blocks – i.e., one hour)

23 .Security Threat !  

24 .Security Threat!   Threat: attacker outruns good miners

25 .Security Threat!   Threat: attacker outruns good miners  Security Assumption : good miners own >.5 of the total compute power [ blockchain.inf o, April 2015] 20%

26 .Permissionless Blockchains Open membership, but inefficient Vulnerable to 50% attacks Examples include Bitcoin , Ethereum , IOTA 26

27 .Permissioned Blockchains Performance: High Throughput, Low Latency Energy-efficient Security: No forks! Closed membership Examples include Ripple, Hyperledger

28 .Blockchain for the Farm? Blockchains require strong network connectivity and lots of storage Permissionless blockchain are power-hungry Sensors have limited resources Sensors for growing conditions, storage conditions, shipping conditions, … Blockchain for a farm will generate records in a decentralized way, and hence it *must* work in a network-partitioned or -challenged environment 28

29 . Vegvisir : tolerate branches Leads to DAG structure instead of linear blockchain Still maintains full causal history of events (respect’s Lamport’s )