2. Problem Statement
• End to end Internet of Things (IoT) applications (I.e. All layers including; sensors,
actuators, network, and cloud) must be continuously monitored to ensure correct
operation.
• Determining when and what to monitor is largely left to the system
administrators, and therefore current monitoring strategies are ad hoc, of
variable quality, and not scalable to large systems.
• Current strategies monitor a single layer in isolation from other layers, thus
solutions are agnostic to policies across dependent IoT components.
• This leads to incomplete and inconsistent monitoring solutions, which is of critical
importance to applications such as health care.
• The aim of this research project is to investigate comprehensive and automated
solutions for end-to-end IoT application monitoring that take into consideration
the policies of all layers and components of an IoT eco-system.
2
8. P2P features
• (L, E Keong, et al. 2005) point out that:
• Selection of nearby peers
• Redundant storage
• Data permanence or guarantee
• Hierarchical naming (L. GARCÉS-ERICE, et al, 2003)
• Trust and authentication
• Anonymity
• Optimized cost and resource utilization.
• Peers can share computing resources without [total] dependency on a central
cloud or server (third party). (P B et al., Empowering the Edge)
8
10. QoS-Aware Deployment of IoT Applications
Through the Fog
• Proposed by (A. Brogi and S. Forti, 2017)
• QoS assurance concerning application deployment over the fog.
• Observation: The majority of tools for automated deployment of
distributed software do not deal with nonfunctional properties to achieve
eligible deployments over available fog nodes.
• Specifically, it concerns segmentation of applications (components).
• Among the possible QoS metrics, they examine latency and bandwidth since nothing
can be done to tame them at the run-time.
• It only proposes best deployment suggestions (environmentally adaptive).
• The select of which particular deployment of a preference is not automated. This is
left to the experts to decide.
10
11. QoS-Aware approach to monitor violations of
SLAs in the IoT (M. Alodib, 2016)
• Discrete EVENT System (DES) to automate generation of QoS awareness and real-time
monitoring.
• A method, which automates the modeling of SLAs, and generates a real-time monitoring service.
• Model Driven Architecture is employed to automate the process of generating a
monitoring service.
• Motivation:
• The absence of standardized mechanisms to represent these requirements.
• The difficulty of declaring SLAs that adapt to dynamic environment.
• Discrete Event System is a diagnostic theory proposed to discover failure.
• Quality of Service (QoS) requirements are declared via Service Level Agreements (SLAs).
• The SoA composite application and user-defined SLAs are mapped into Petri net model.
• Therefor, An algorithm to compute and produce a QoS-Aware service.
• QoS-aware service monitors any violations of SLAs.
11
12. IoT Monitoring Challenges
• Most of the current SLA monitoring tools are provider-dependent(T. Labidi, et al, 2017).
• Heterogeneity
• Data format, communication protocols, OS, policies, batch production, etc.
• Agnostic environment.
• Modern large scale applications are not monolithic anymore.
• An application running over a Fog computing infrastructure can be thought as a set of
independently deployable components that are working together and must meet some QoS
constraints.
• Formal definition of the performance metrics (E. Solaiman et al. 2016)
• Each provider would suggest its own SLA definition, thresholds and provider-dependent
equations.
• Preventing breach occurrence at the first place. Especially to critical systems
• It is of importance to ensure correct operation in advance.
• Distributed nature of IoT complicates the task of breach monitoring (M. Alodib, 2016).
12
16. Smart contracts
• Ethereum (GAVIN WOOD, 2014)
• Ether is the digital fuel that motivates the Ethereum network
• Turing-complete programming
• Solidity https://solidity.readthedocs.io/en/develop/
• “Arbitrary contracts can theoretically be created for any transaction type or application”.
(Ethereum: white papre)
• Business logic is expressed as a code (K. Christidis et al, 2016)
• Deterministic behavior
• Same input -> same output
• Smart contracts allows the following:
• Maintaining direct realtime control.
• Governance rules are formalized.
• Contracts are self-enforced and executed.
16
18. Consensus
• Common truth is called consensus among nodes.
• The process of keeping the ledger transactions synchronized across
the network.
• All appropriate participants agree on the transaction status.
• Consensus has to be Byzantine Fault-Tolerant using the PBFT protocol
and others means depending on the scenario (C. Cachin, 2016).
• Consensus mechanism selection is influenced by whether the
network is private (permissioned) or public (permissionless)(M. Valenta and P.
Sandner, 2017)
18
23. • Basically, they use blockchain to store events such as
• PeriodicPayout , InsufficientThroughput, throughputBreach etc.
• Event can be locked up and referred to.
• Once a MONITOR in either parties observes a violation,
• It specifies the the breach with respect to the QoS.
• The contract is triggered.
• This contract then will apply penalty.
23
24. • Observations by (E. Di Pascale et al., 2017)
• Advantages:
• The use of smart contracts can automatically impose rewards and penalty mechanisms.
• Complex business models can be easily transformed into automatically enforceable
contracts.
• Limitations:
• Legality issues: no central authority able to rectify a transaction over a permissionless
ledger in case of a dispute.
• Smart contracts: might be vulnerable to bugs and attacks.
• Recommendation:
• It is recommended to include a fail-safe mechanism in the contract to be able to disable
it and recover any outstanding balance in its possession in the event of the discovery of a
vulnerability in the code.
24
25. Blockchain limitations
• The performance is a cost for trustless communication.
• lower transaction processing throughput and higher latencies.
• This is mainly due to proof-of- work mechanisms. (M. Vukolić, 2015)
• A block is created within ~10 min.
• Byzantine failure.
• Latency: due to the fault-tolerant gossip process(broadcasting) (C. Decker and R. Wattenhofer, 2015).
• Scalability
• Centralized platform: Visa can process ~47,000 transactions/s.
• Blockchains: Depending on block size limit, say 1 MB: in Bitcoin -> ~ 7 transactions/s.
• synchronized global state, the replicated blockchain hinder scalability
• How to scale blockchains Without sacrificing the decentralization and security. (J. Poon et al., 2016).
• Unconfirmed transactions https://blockchain.info/unconfirmed-transactions
• Security
• The fact of smart contracts represented as a code, it can be vulnerable to attacks (E. Di Pascale et al., 2017)
.
• Review this more detail https://www.slideshare.net/_hd/ibm-adept
25
28. ADEPT
• The paper suggests that, a decentralized IoT must support:
• Trustless, encrypted messaging and transport
• Low latency with guaranteed delivery
• The concept of Distributed Hash Tables (DHTs) can meet requirements
• Telehash is best open source messaging implementation. Thereby, it is used.
• Basically, Notifications among devices without using a centralized server.
• Bit Torrent DHT file sharing protocol
• Used for propagating purposes and content distribution. (No centralization)
28
35. • Tangle is a Distributed Ledger for M2M Payments and interactions:
• Automated execution and security.
• It is based on the concept of Directed Acyclic Graph (DAG).
• There is nether blocks, nor chains.
• No Miners -> Zero-incentive.
• punishment strategy instead of incentives.
• All active participants are directly involved.
• It shares with blockchain essential principals but not necessary the
underlying mechanism:
• Distributed storage and P2P network.
• Consensus and validation.
35
37. • Consensus is no longer decoupled from transaction processes, which
enables a validation without fees.
• A new transaction attests directly other two transactions, and indirectly that a
subsection of the Tangle are valid and conforms to the protocols rules.
• The transaction is signed with private key
• MCMC (Markov chain Monte Carlo) randomly selects existing unconfirmed
transactions (tips).
• Proof of work similar to Hashcash (spam and sybil-resistance).
• New transactions will try to confirm this transaction with same procedure.
• They claim this consensus mechanism allows micropayment with
scalability.
• Consensus is parallelized and is not sequential. Fast
• Test: 250 nodes -> ~ 100 confirmed transaction/s.
37
39. The Bitcoin Lightning Network: Scalable Off-Chain
Instant Payments (J. Poon and T. Dryja, 2016)
• This paper proposes a network of micropayment channels.
• For blockchains to deal with large volume of data and a massive
number of transactions, there has to be a solution that does not
compromise the decentralisation. Otherwise, it is either the
blockchain collapses or the more centralised approach will replace.
• Brining the validation process close to the consumer level.
• So instead of mining every transaction, only mining the channel
opening/closing. https://www.slideshare.net/tm89lee/intro-to-lightning-network-bitcoinlitecoin-blockchain-developers-Malaysia.
• Micropayment channels only create a relationship between two parties.
• Transactions are instantly executed off-chain. No miners involved.
https://www.slideshare.net/meinhard/bitcoin-micropayment-channels
39
42. Architecture of the Hyperledger Blockchain
Fabric (C. Cachin, 2016)
• Hyperledger project started in December 2015 by the Linux Foundation.
• Several major blockchain initiatives composing the heyperLedger on it including IBM,
Intel, .. and other 50+ members, Hence the name.
• cross-industry blockchain technologies. Standardization
• Open-source Modular Architecture.
• Pluggable implementations of different components.
• It promises high degrees of confidentiality, resiliency, flexibility, scalability.
• performance is achieved due to the obviation of PoW and permissioned
environment.
• Hyperledger Fabric is private and permissioned.
• Membership Service Provider (MSP) instead of Proof of Work.
• determine who participate in the validation process
42
46. Highlights & Remarks
• Regarding the IoT monitoring with blockchain technology:
• Why blockchain at the first place?
• What should be recorded on the ledger and why?
• Should we have a global ledger or community-based ledger?
• Is smart contracts expression-rich to cover any legal clauses and terms?
• How about the dynamic changes in terms of requirements ( sustainability)?
• What best blockchain architecture and alternatives we should adopt?
• Is blockchain really scalable enough to handle IoT massive interactions?
• Either a unified SLAs formulation or an approach to deal with this
heterogeneity?
• How to scale blockchains Without sacrificing the decentralization and
security?
46
47. Conclusion
• The need to address scalability of blockchain.
• The solution should be incentive free.
• Smart contract can model governance rules.
• A fail-safe mechanism should be introduced into smart contracts.
• In our case, there is no need for strict PoW/PoS.
• Fog/edge computing should be essential to any introduced solution.
• We need stress the significance of decentralized computation.
• Instead of validating every single transaction, We could only validate
channels among things.
• HyperLedger Fabric sounds promising.
47
48. References
• Cisco, “The Internet of Things Reference Model,” 2014.
• P. Brody and V. Pureswaran, ‘‘Device democracy: Saving the future of the Internet of Things,’’ IBM Institute for Business Value, Tech. Rep., Sep. 2014.
• L. GARCÉS-ERICE, E. W. BIERSACK, K. W. ROSS, P. A. FELBER, and G. URVOY-KELLER, “HIERARCHICAL PEER-TO-PEER SYSTEMS,” Parallel Process. Lett., vol. 13, no. 4, pp. 643–657, 2003.
• J. Poon and T. Dryja, “The Bitcoin Lightning Network: Scalable Off-Chain Instant Payments,” 2016.
• M. Chiang and T. Zhang, “Fog and IoT: An Overview of Research Opportunities,” IEEE Internet Things J., vol. 3, no. 6, pp. 854–864, Dec. 2016.
• T. Labidi, A. Mtibaa, W. Gaaloul, S. Tata, and F. Gargouri, “Cloud SLA Modeling and Monitoring,” in 2017 IEEE International Conference on Services Computing (SCC), 2017, pp. 338–345.
• E. Solaiman, R. Ranjan, P. P. Jayaraman, and K. Mitra, “Monitoring Internet of Things Application Ecosystems for Failure,” IT Prof., vol. 18, no. 5, pp. 8–10, Sep. 2016.
• M. Vukolić, ‘‘The quest for scalable blockchain fabric: Proof-of-work vs. BFT replication,’’ in Proc. IFIP WG 11.4 Workshop Open Res. Prob- lems Netw. Secur. (iNetSec), 2015, pp. 112–125.
• K. Christidis and M. Devetsikiotis, “Blockchains and Smart Contracts for the Internet of Things,” IEEE Access, vol. 4, pp. 2292–2303, 2016.
• Y. Sompolinsky and A. Zohar, “Secure High-Rate Transaction Processing in Bitcoin,” Springer, Berlin, Heidelberg, 2015, pp. 507–527.
• “Ethereum: White Paper.”, https://github.com/ethereum/wiki/wiki/White-Paper.
• GAVIN WOOD, “ETHEREUM: A SECURE DECENTRALISED GENERALISED TRANSACTION LEDGER,” 2014.
• P. B. Veena Pureswaran, Sanjay Panikkar, Sumabala Nair, “Empowering the edge,” IBM Inst. Bus. Value. https://www-935.ibm.com/services/multimedia/GBE03662USEN.pdf
• Lua, Eng Keong, et al. "A survey and comparison of peer-to-peer overlay network schemes." IEEE Communications Surveys & Tutorials 7.2 (2005): 72-93.
• V. P. Sanjay Panikkar, Sumabala Nair, Paul Brody, “ADEPT: An IoT Practitioner Perspective,” IBM Inst. Bus. Value, 2015.
• J. Poon and T. Dryja, “The Bitcoin Lightning Network: Scalable Off-Chain Instant Payments,” 2016.
• S. Popov, “The Tangle,” 2017.
• https://blog.iota.org/a-primer-on-iota-with-presentation-e0a6eb2cc621
• C. Jentzsch, “Decentralized Autonomous Organization to Automate Governance,” SlockIt, pp. 1–30, 2016.
• A. Brogi and S. Forti, “QoS-aware Deployment of IoT Applications Through the Fog,” IEEE Intfile//campus/home/home2016/b6072750/Desktop/Articles/QoS-aware Deploy. IoT Appl. Through Fog.pdfernet Things J., pp. 1–1, 2017.
• R. Milito, F. Bonomi, R. Milito, P. Natarajan, and J. Zhu, “Big Data and Internet of Things: A Roadmap for Smart Environments,” vol. 546, no. January, 2014.
• M. Alodib, “QoS-Aware approach to monitor violations of SLAs in the IoT,” J. Innov. Digit. Ecosyst., vol. 3, no. 2, pp. 197–207, 2016.
• E. Di Pascale, J. McMenamy, I. Macaluso, and L. Doyle, “Smart Contract SLAs for Dense Small-Cell-as-a-Service,” no. May, pp. 2–4, 2017.
• C. Decker and R. Wattenhofer, “A Fast and Scalable Payment Network with Bitcoin Duplex Micropayment Channels,” in Stabilization, Safety, and Security of Distributed Systems: 17th International Symposium, SSS 2015, Edmonton, AB, Canada, August 18-21, 2015, Proceedings, A.
Pelc and A. A. Schwarzmann, Eds. Cham: Springer International Publishing, 2015, pp. 3–18.
• H. Nakashima and M. Aoyama, “An Automation Method of SLA Contract of Web APIs and Its Platform Basedon Blockchain Concept,” Proc. - 2017 IEEE 1st Int. Conf. Cogn. Comput. ICCC 2017, pp. 32–39, Jun. 2017
48