Personal and Ubiquitous Computing
https://doi.org/10.1007/s00779-019-01339-5
ORIGINAL ARTICLE
MBS: Multilevel Blockchain System for IoT
Bacem Mbarek1 · Nafaˆa Jabeur2 · Tom´as Pitner1 · Ansar-Ul-Haque Yasar3
Received: 19 August 2019 / Accepted: 18 October 2019
© Springer-Verlag London Ltd., part of Springer Nature 2019
Abstract
Despite of their increasing popularity, Internet of Things (IoT) platforms are still suffering from major security problems,
particularly during communications between the IoT devices. Indeed, these devices are commonly prone to malicious
attacks and require prior mutual authentication to guarantee the confidentiality and security of data being shared as well
as a proper network operation. In order to deal with this issue, several researchers have proposed the use of the emergent
blockchain paradigm, which has emerged as a key technology that will transform the way in which information will be
shared. As this paradigm is facing scalability and flexibility problems, the use of multi-agents systems has been adopted in
recent works. However, current solutions are using agents for the rudimentary role of data collection only. In this paper, we
propose to secure the IoT platform with a multilevel blockchain system (MBS) where the speed and flexibility of blockchain
transactions are enforced by mobile agents which are migrating throughout the IoT network. The simulations of our solution
through the Hyperledger Fabric are showing relevant results in terms of response time and energy consumption.
Keywords Blockchain · Internet of Things · Security · Scalability
1 Introduction
Blockchain has recently emerged as a promising technology
for sharing and protecting vulnerable data in distributed
and decentralized systems [1]. It is basically using publickey
cryptography to sign transactions between parties.
Transactions are then stored in a distributed ledger. The
ledger consists of cryptographically linked blocks of data
(i.e., transactions), which are almost impossible to change
Bacem Mbarek
bacem.mbarek1@gmail.com
Nafaˆa Jabeur
nafaa.jabeur@gutech.edu.om
Tom´as Pitner
tomp@fi.muni.cz
Ansar-Ul-Haque Yasar
ansar.yasar@uhasselt.be
1 Faculty of Informatics, Masaryk University,
Brno, Czech Republic
2 German University of Technology in Oman (GUtech),
Athaibah, Sultanate of Oman
3 Transportation Research Institute Hasselt University,
Hasselt, Belgium
or remove. Thanks to their enforced security, blockchain
technologies (BCT) are attracting an increasing interests
from the industrial, academic, and governmental sectors.
They have been already used in several application fields,
such as banking, asset management, smart appliances, and
healthcare. Within the context of Internet of Things (IoT),
BCT are being used to address various challenges, including
decentralization, anonymity, and trust [2–4]. However, the
performance of the proposed solutions still needs to be
improved, particularly since blockchain is fundamentally
computationally expensive and involves high bandwidth
overhead and delays and its related technologies are still
immature. As the IoT devices are becoming capable of
acting and reacting to events in their environments, we
argue that blockchain solutions could benefit from the
collaboration capabilities of these devices to enhance
the processes of authentication and coordination of data
collection as well as optimizing the use of the available
resources.
Collaboration in the IoT filed is basically allowing the
IoT devices to achieve tasks exceeding their individual
capabilities. Because of the commonly limited processing,
storage, and communication capabilities of these devices as
well as the dynamic, open, and unpredictable environments
in which they operate, implementing effective collaboration
mechanisms remains a challenging concern. Potential
solutions could be found/inspired from the extensive studies
Pers Ubiquit Comput
on enabling collaboration between distributed autonomous
entities done in the context of multi-agent systems (MAS).
MAS have, indeed, proven flexibility, autonomy, and intelligence
to solve complex problems within highly dynamic,
constrained, and uncertain environments [5, 6]. In this
regard, proposed approaches have focused on endowing distributed
software agents with intelligence to autonomously
and proactively make decisions based on current activities,
contextual information, and the envisioned applications.
These agents have been successfully used, for example,
to exchange sensitive data [7], plan critical missions (e.g.,
[8, 9]), optimize energy distributions (e.g., [10, 11]), and
improve healthcare services (e.g., [12, 13]).
Being motivated by this success, recent works (e.g., [14–
17]) have highlighted the relevance of integrating MAS
and BCT. On the one hand, BCT could solve the security
limitations of MAS, thanks to their capabilities of dealing
with decentralization, anonymity, and trust issues [18]. On
the other hand, MAS can highly improve the flexibility and
scalability of blockchain solutions [19], which are not yet
capable of efficiently fulfilling the requests of high volumes
of concurrent access to blockchain platforms [18]. Mobile
agents can also authenticate, optimize, and coordinate the
data collection process for blockchain activities. In spite
of the inspiring ideas presented in the works combining
the use of MAS and BCT, we argue that additional efforts
are still needed to allow for a seamless integration of
both paradigms. Indeed, in the particular field of IoT
applications, the proposed MAS architectures have limited
configurability, scalability, and efficiency. In addition, the
current use of mobile agents is limited to the rudimentary
role of migrating to IoT devices to bring back data to the
blockchain platform. The coordination of actions between
the different levels of the IoT network is also not being
supported correctly. We further consider that our proposed
solution is application agnostic and well suited for various
IoT applications. That means any use case is feasible where
the blockchain collects sensor data.
We, therefore, propose in this paper a new secure, private,
and lightweight multilevel blockchain system (MBS) for
IoT applications. Our solution is based on the use of mobile
agents to reduce blockchain processing and communication
overheads. To this end, the agents collaboratively execute
blockchain functions at different levels of the IoT network
before transferring the hashed data blocks they create to the
blockchain platform. Compared to the existing solutions,
our contributions could be summarized as follows: (1) a new
multilevel architecture combining the three paradigms of
IoT, blockchain, and multi-agent systems; (2) a new solution
enabling the improvement of the blockchain scalability
via the use of mobile agents that migrate throughout the
IoT network and perform blockchain activities at different
hierarchical levels of this network; and (3) a new multilevel
solution based on mobile agents to reduce energy
consumption and communication overheads in blockchainIoT
applications.
In the remainder of this paper, Section 2 outlines the
existing works that have addressed the use of BCT within
the IoT context. It also outlines the use of MAS along
with BCT in order to improve IoT applications. Section 3
describes our proposed solution. It particularly focuses on
our proposed authentication scheme as well as on our use of
BCT and mobile agents toward an improved authentication
process. Section 4 provides a performance evaluation of our
solution. Section 5 summarizes the paper and highlights our
future works.
2 Related work
Because of the commonly restricted resources and the heterogeneity
of their devices, IoT networks are facing serious
security problems [20]. These problems are particularly significant
due to the increasing amounts of data which are
being collected and shared between the IoT devices [21].
In order to overcome these problems, recent works (e.g.,
[22, 23]) have highlighted the relevance of using the emergent
BCT along with the IoT paradigm. In this regard,
Jesus et al. [24] have presented a survey on how to use
blockchain to secure the IoT network and its related activities.
Golomb et al. [25] have proposed the solution CIoTA
where blockchain is used jointly with extensible Markov
model (EMM) to discover abnormalities in IoT applications.
Dorri et al. [21] have investigated the breaches in existing
security and privacy methods and proposed a lightweight
and scalable blockchain (LSB) solution to reduce bandwidth
and computation costs while improving IoT security
and privacy. Huckle et al. [26] have investigated the contributions
of blockchain for monetizing IoT applications.
Huh et al. [27] have demonstrated how blockchain can
be used to store IoT devices’ data with smart contracts.
Shafagh et al. [28] have proposed a new framework where
blockchain is used to support the distributed storage of
data in IoT applications. Danzi et al. [29] have proposed
various mechanisms/protocols that enable the synchronization
of IoT endpoints to the blockchain. To this end, the
authors have used several security levels and communication
requirements. They have also analyzed the traffic and
the power consumption created by the synchronization protocols
via numerical simulations. Extended reviews of using
Pers Ubiquit Comput
blockchain to store IoT data (with a focus on open problems
of anonymity, integrity, and adaptability) as well as to architect
blockchain-based IoT applications are available in [30]
and [31] respectively.
In spite of their relevance, the proposed solutions
are suffering from the problems of scalability, delays,
and bandwidth overhead inherited from the blockchain
paradigm. As demonstrated by Danzi et al. [29], these
problems could get worse due to high communication
exchanges between the IoT devices. Furthermore, reaching
consensus, for example on adding a new block to
the ledger, can be extremely complex in a distributed
network (like the IoT) where peers may have unaligned
interests [32]. To overcome such issues, recent works (e.g.,
[33, 34]) have supported the idea of integrating multiagent
systems (MAS) with BCT. In this context, Ferrer
[35] has demonstrated that by integrating cryptographic
algorithms with peer-to-peer (P2P) networks, software
agents can autonomously be able to find consensus on
particular state of affairs. Calvaresi et al. [18] have
highlighted that MAS combined with BCT can provide
the necessary means to make the operations of distributed
entities more autonomous, flexible, secure, and profitable.
Kapitonov et al. [19] have presented a solution to organize
communications involving software agents in an IoT
network. The solution uses the decentralized Hyperledger
blockchain technology as well as smart contracts. In this
paper, we argue that the potential of software agents,
and particularly mobile agents, is still untapped. We also
argue that the right support from agents will definitely
leverage the performance of any combined blockchain-IoT
solution.
3 Multilevel blockchain system (MBS)
3.1 Network model
In order to support the exponential growth of IoT, we
are proposing a blockchain-based system to enforce the
privacy and security matters related to collecting, sharing,
and managing vulnerable data. To reduce blockchain
overheads while addressing its scalability problems, we are
deploying mobile agents that migrate throughout the IoT
network to collect relevant data, create hashed data blocks,
reduce response time delays, and solve synchronization
and scalability problems. We are summarizing our solution
into a multilevel blockchain system (MBS) that allows
smart devices to securely send their data to the central IoT
processing platform (also blockchain platform). The levels
of the proposed solution are (see Fig. 1): (1) micro-level
(IoT device level); (2) meso-level blockchain (cluster heads
level of the IoT network); and (3) macro-level blockchain
(platform level or also the highest hierarchical level of the
IoT network).
3.2 Overview of our layered design
The implemented platform consists of four parts as shown
in Fig. 2: IoT device, ordering service, endorsing peers, and
committing peers.
– IoT device: These devices are collecting and transmitting
data to private blockchain distributed ledgers.
– Endorsing peers: During the commissioning and
configuration of the blockchain network, the developers
Fig. 1 Hierarchical architecture
for data management
Pers Ubiquit Comput
should select a number of IoT devices defined as the
endorsing peers.
– Ordering service: Creates the block of transactions,
and sends it to all the peers. The ordering service
collects transactions for a channel into proposed blocks
for distribution to peers. Blocks are delivered on a
channel basis. The ordering service accepts endorsed
transactions, orders them into a block, and delivers the
blocks to the committing peers. The main responsibility
of ordering service is to receive transactions from the
IoT devices and fit into a block.
– Committing peers: A committing peers are a predefined
number of IoT devices. Usually endorsing peers are
also committing, but a peer can be only committing and
not endorsing. Committing peers (including endorsing
peers) run validation and update their copy of the
blockchain and world state. Each peer receiving the
block, now in the role of a committing peer, appends the
whole block to its blockchain copy. Committing peers
are responsible for adding blocks of transactions to the
shared ledger and updating the world state. They may
hold smart contracts, but it is not a requirement.
3.2.1 Multiagent system model
Our solution includes three types of agents: (1) macro
agents, (2) meso agents, and (3) micro agents. The role of
these agents and their locations are depicted on Fig. 2 and
explained in what follows.
Macro agent: A macro agent is located on the highest
level of the IoT, which is also called the blockchain platform
(BPL). This agent is responsible of authenticating
the IoT clusters and their cluster heads. It is also responsible
of creating the related meso agents and collecting
from them the encrypted blocks of data of interest. The
macro agent will then update the blockchain accordingly
and store related data for subsequent uses within the
context of the envisioned application.
Meso agents: We assume here that the IoT network is
following a specific algorithm to create clusters and elect
their heads (this algorithm is out of the scope of this
paper). Once authenticated at the macro level of the
system, a given cluster will receive its specific meso
agent. This agent will collect and aggregate the data of
interest from the micro agents assigned to the IoT devices
of its cluster. It will then create and encrypt the data block
corresponding to its cluster. The meso agent will then
migrate to the macro level to deliver the data to the macro
agent.
Micro agents: The micro-level of the IoT hierarchy is
composed of IoT devices. Once an IoT device is
authenticated as member of a given cluster, its related
micro agent will be created by the meso agent of the
Fig. 2 Order-execute architecture to store data in MBS blockchain platform
Pers Ubiquit Comput
cluster. The micro agent will migrate to the IoT device
and collect the data of interest periodically or on demand.
The agent will then encrypt the data and migrate back to
the meso level (i.e., the head of its cluster).
3.2.2 Order-execute architecture in MBS blockchain
platform
Figure 2 shows the order-execute architecture to data
storage in the MBS blockchain platform. This architecture
can be summarized as follows.
1. The IoT device submits endorsement requests to
endorsing peers.
2. Endorsing peers approve or reject the status after
checking for consistency (checking the smart con-
tract).
3. Endorsing peers send the proposal response to the
IoT device applicant. Each execution captures the set
of read and written data (also called the RW set),
which is flowing in the Hyperledger Fabric. Moreover,
the endorsement system chaincode (ESCC) signs the
proposal response on the endorsing peer . The RW sets
are signed by each endorser. Every set will include the
number of the version of each record.
4. Afterwards, the ordering service located in the micro
agent verifies the collected endorsements and creates
the block of transactions.
5. The micro agent sends the created block to the mesolevel
blockchain (cluster heads level).
6. The meso level updates the received block and adds
transactions received from different IoT devices.
7. Then, the meso agent sends the block to all the peers of
the network to be appended to their blockchain copies.
8. The peers check the transaction’s validity (verification
of the smart contract).
9. If the validity is fulfilled, finally, the macro level
(blockchain platform) adds the checked block into the
data base of the blockchain platform.
10. Peers emit notification.
3.2.3 Smart contracts for monetizing IoT data
The concept of smart contracts can be used to automate
negotiations between service providers and users along
with required monetary transactions without a trusted
intermediary. A smart contract is fundamentally a code that
validates a negotiation and immediately brings a contract
into effect, without the involvement of any intermediaries
[36]. The smart contract code resides on a blockchain as
multiple functions with unique addresses that can be called
by any user of the blockchain. All entities interacting with a
blockchain (including users and Things) must own at least
one public-private key pair. First of all, the sender encrypts
a smart contract with the public key. The receiver then
receives the encrypted smart contract and decrypts it with
the private key.
4 Performance evaluation
In order to implement our multilevel blockchain system
(MBS), we used Hyperledger Fabric [37], which is an open
source permissioned blockchain introduced by IBM. We run
our simulations on a large-scale network of 1000 nodes.
We assumed that all nodes have a fixed position throughout
the period of simulations, with the settings provided in
Table 1. The results have been obtained by a confidence
interval of 95%. We compared our MBS solution with the
solution proposed in [19], which focuses on organizing a
communication system between agents (representing smart
things or robots) in a P2P network using the decentralized
Hyperledger blockchain technology and smart contracts to
reduce energy consumption. This choice was motivated
by the fact that both solutions are integrating the use of
MAS and blockchain at several levels of the IoT hierarchy.
The setup of our experiments is as follows: (1) nodes
run on Fabric version v1.1.0-preview2 instrumented for
performance evaluation through local logging; (2) all nodes
are 2.0 GHz 16-vCPU VMs running Ubuntu with 8 GB of
RAM and SSDs as local disks; (3) there are 1000 peers
in total, 100 IoT devices endorsers; (4) signatures use the
default SHA-1 scheme; and (5) we adopt 2 MB as data block
sizes. During our experiments, we analyzed the size mint
and spend transactions. In particular, the 2-MB data blocks
contained 473 mint or 670 spend transactions. In other
words, the average transaction size is 3.06 kB for spend and
4.33 kB for mint. In general, transactions in Fabric are large
because they carry certificate information.
We have implemented a micro agent in each IoT device
as a Hyperledger Fabric ordering service to create blocks.
Micro agents’ actions are executed when an IoT device
received and verified a smart contract request. Each node in
the system runs Ubuntu with 8GB of memory. We have also
implemented a meso agent in each cluster head which can
Table 1 Simulation parameters
Parameter Value
Endorsing peers 100 peers
Cluster head 50
Block size 2MB
Run times 50 times
Simulation time 100 s
Number of nodes 1000
Pers Ubiquit Comput
collect the received blocks from different IoT devices. We
install on the blockchain platform a macro agent to validate
the chain of blocks received from the cluster head and to
update the blockchain platform.
4.1 Response time
Figure 3 outlines our assessment of response time for both
solutions MBS and BCE. The figure clearly demonstrates
that our solution outperforms the solution proposed by [19].
For instance, for the same size of network consisting in
1200 nodes, the response time of BCE is around 9 ms,
compared to about 5 ms in MBS. This could be explained
by the fact that BCE uses a single blockchain level installed
in a remote server platform. In this configuration, decisions
about the necessary actions to perform (e.g., data hashing,
data encryption) will be delayed as many mobile agents will
not be able to concurrently access the blockchain platform
to insert their data. In our solution, mobile agents at different
levels of the IoT network have already the rights and the
duty to perform blockchain actions ahead before migrating
to the remote blockchain platform. Once there, they will not
be constrained to wait until blockchain actions are executed
and relevant actions are taken accordingly.
4.2 Energy consumption
Figure 4 gives the average energy consumption in Joules
which is needed to store the data in the blockchain
platform. Compared to the BCE where each data transaction
is processed and sent individually, our MBS solution
consumes less energy since our meso agents are collecting
Fig. 3 Average response time to store data in blockchain platform
Fig. 4 Energy consumption
and aggregating blocks of data to reduce communication
overheads. For example, as could be seen on Fig. 4, at time
of 8 hours, BCE uses about 250 (j) whereas MBS consumes
only 190 (j).
4.3 Use case: Internet of bikes
The IoB (Internet of Bikes) is one of the widely spreading
examples of distributed applications [38], urging for
highly secure, independent, and distributed platform, which
blockchain is capable of supporting. Moreover, blockchain
needs to be established in IoB applications in order to truly
leverage the benefits provided by its distributed ledger to
find solutions to problems in bike sharing such as bike
deployment, redistribution, parking, and maintenance. Due
to the nature of the blockchain, all bikes information in
the edge can be tracked in a synchronized and distributed
blocks.
In order to prevent such threats, we are proposing to
send encrypted bike information based on blockchain. In
addition, for guaranteeing confidentiality, our proposed
security scheme allows for encrypting and generating
illegible messages. It has been proven (e.g., [39]) that
these messages are only accessible by authenticated parties.
For additional confidentiality, smart contract is used
alongside with cryptography. Therefore, we are proposing
a blockchain-based system to enforce the privacy and
security matters related to collecting, sharing, and managing
vulnerable data. Moreover, we encounter the blockchain
overhead communication issues by deploying agent that
collects data, creates encrypted data, creates blocks, and
Pers Ubiquit Comput
solves synchronization and scalability problems in IoB
environment.
5 Conclusion
We proposed in this paper a multilevel blockchain system
(MBS) to enhance current data security and privacy in
IoT applications while improving response time and energy
consumption. Our solution is based on the use of mobile
agents that are capable of migrating between the different
levels of the IoT network and the blockchain platform
and execute the necessary blockchain functions as well as
hashing, encryption, aggregation, and decryption functions
when and where needed. Simulations of our solutions are
showing satisfactory results. These results are motivating us
to focus our future works on implementing our solution in a
real IoT case study.
References
1. Back A, Corallo M, Dashjr L, Friedenbach M,
Maxwell G, Miller A, Poelstra A, Tim´on J, Wuille P
Enabling blockchain innovations with pegged sidechains,
http://www.opensciencereview.com/papers/123/enablingblockchain-
innovations-with-pegged-sidechains
2. Li Z, Kang J, Yu R, Ye D, Deng Q, Zhang Y (2018) Consortium
blockchain for secure energy trading in industrial internet of
things, IEEE Transactions on Industrial Informatics
3. Signorini M, Pontecorvi M, Kanoun W, Di Pietro R (2018) Bad:
blockchain anomaly detection. arXiv:1807.03833
4. Mbarek B, Jabeur N et al (2018) Ecass: an encryption compression
aggregation security scheme for secure data transmission in ambient
assisted living systems. Personal and Ubiquitous Computing
5. Wray K, Thompson B (2014) An application of multiagent
learning in highly dynamic environments. In: AAAI workshop on
multiagent interaction without prior coordination
6. Kong Y, Zhang M, Ye D (2017) A belief propagationbased
method for task allocation in open and dynamic cloud
environments, Knowledge-Based Systems
7. Schatten M, ˇSeva J., Tomiˇci´c I. (2016) A roadmap for scalable
agent organizations in the internet of everything, Journal of
Systems and Software
8. Wong D, Paciorek N, Moore D (1999) Java-based mobile agents.
Commun ACM 42(3):92–ff
9. Caripe W, Cybenko G, Moizumi K, Gray R (1998) Network
awareness and mobile agent systems. IEEE Commun Mag
36(7):44–49
10. Tong L, Zhao Q, Adireddy S (2003) Sensor networks with
mobile agents. In: Military communications conference, 2003.
MILCOM’03. 2003 IEEE, vol 1. IEEE, pp 688–693
11. Dong M, Ota K, Yang LT, Chang S, Zhu H, Zhou Z (2014)
Mobile agent-based energy-aware and user-centric data collection
in wireless sensor networks. Comput Netw 74:58–70
12. Su C-J, Wu C-Y (2011) Jade implemented mobile multi-agent
based, distributed information platform for pervasive health care
monitoring. Appl Soft Comput 11(1):315–325
13. Chan V, Ray P, Parameswaran N (2008) Mobile e-health
monitoring: an agent-based approach, IET communications
14. Kvaternik K, Laszka A, Walker M, Schmidt D, Sturm M, Dubey A
et al (2017) Privacy-preserving platform for transactive energy
systems. arXiv:1709.09597
15. Qayumi K (2015) Multi-agent based intelligence generation from
very large datasets. In: IEEE international conference
16. Norta A, Othman AB, Taveter K (2015) Conflict-resolution
lifecycles for governed decentralized autonomous organization
collaboration. In: Proceedings of the 2015 2nd international
conference on electronic governance and open society: Challenges
in Eurasia. ACM, pp 244–257
17. Ponomarev S, Voronkov A (2017) Multi-agent systems and
decentralized artificial superintelligence. arXiv:1702.08529
18. Calvaresi D, Dubovitskaya A, Calbimonte JP, Taveter K,
Schumacher M (2018) Multi-agent systems and blockchain:
results from a systematic literature review. In: International
conference on practical applications of agents and multi-agent
systems. Springer
19. Kapitonov A, Lonshakov S, Krupenkin A, Berman I (2017)
Blockchain-based protocol of autonomous business activity for
multi-agent systems consisting of uavs. In: Research education
and development of unmanned aerial systems (RED-UAS)
20. Jing Q, Vasilakos AV, Wan J, Lu J, Qiu D (2014) Security of
the internet of things: perspectives and challenges. Wireless Netw
20(6):2481–2501
21. Dorri A, Kanhere SS, Jurdak R, Gauravaram P (2017) Lsb:
A lightweight scalable blockchain for iot security and privacy.
arXiv:1712.02969
22. Kshetri N (2017) Can blockchain strengthen the internet of things?
IT Professional 19(4):68–72
23. Christidis K, Devetsikiotis M (2016) Blockchains and smart
contracts for the internet of things. IEEE Access v4:2292–2303
24. Jesus EF, Chicarino VR, de Albuquerque CV, Rocha A. A. d. A.
(2018) A survey of how to use blockchain to secure internet
of things and the stalker attack, Security and Communication
Networks
25. Golomb T, Mirsky Y, Elovici Y (2018) Ciota:, Collaborative iot
anomaly detection via blockchain. arXiv:1803.03807
26. Huckle S, Bhattacharya R, White M, Beloff N (2016) Internet
of things, blockchain and shared economy applications. Procedia
Comput Sci 98:461–466
27. Huh S, Cho S, Kim S (2017) Managing iot devices using blockchain
platform. In: 2017 19th international conference on advanced
communication technology (ICACT). IEEE, pp 464–467
28. Shafagh H, Burkhalter L, Hithnawi A, Duquennoy S (2017)
Towards blockchain-based auditable storage and sharing of iot
data, In: Proceedings of the 2017 on cloud computing security
workshop. ACM, pp 45–50
29. Danzi P, Kalor AE, Stefanovic C, Popovski P (2018) Analysis
of the communication traffic for blockchain synchronization
of iot devices. In: 2018 IEEE international conference on
communications (ICC). IEEE, pp 1–7
30. Conoscenti M, Vetro A, De Martin JC (2016) Blockchain for
the internet of things: a systematic literature review. In: 2016
IEEE/ACS 13th international conference of computer systems and
applications (AICCSA). IEEE, pp 1–6
31. Ramachandran GS, Radhakrishnan R, Krishnamachari B (2018)
Towards a decentralized data marketplace for smart cities. In:
invited paper at the 1st international workshop on BLockchain
enabled sustainable smart cities (BLESS 2018), Kansas City, MO,
USA, held in conjunction with the 4th IEEE Annual International
Smart Cities Conference (ISC2)
Pers Ubiquit Comput
32. Shermin V (2017) Disrupting governance with blockchains and
smart contracts. Strateg Chang 26(5):499–509
33. Kvaternik K, Laszka A, Walker M, Schmidt D, Sturm M, Dubey A
et al (2017) Privacy-preserving platform for transactive energy
systems. arXiv:1709.09597
34. Ponomarev S, Voronkov A (2017) Multi-agent systems and
decentralized artificial superintelligence. arXiv:1702.08529
35. Ferrer EC (2018) The blockchain: a new framework for robotic
swarm systems. In: Proceedings of the future technologies
conference. Springer, pp 1037–1058
36. Di Pierro M (2017) What is the blockchain? Comput Sci Eng
19(5):92–95
37. Hyperledger fabric. https://github.com/hyperledger/fabric
38. Behrendt F (2016) Why cycling matters for smart cities. internet of
bicycles for intelligent transport. Journal of Transport Geography
56:157–164
39. Hamid N, Yahya A, Ahmad RB, Al-Qershi OM (2012) Image
steganography techniques: an overview. Int J Comput Sci Secur
(IJCSS) 6(3):168–187
Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.