Content uploaded by Will Serrano
Author content
All content in this area was uploaded by Will Serrano on Apr 01, 2021
Content may be subject to copyright.
5G Cybersecurity based on the Blockchain Random
Neural Network in Intelligent Buildings
Will Serrano
Intelligent Systems and Networks Group
Electrical and Electronic Engineering
Alumni Imperial College London
g.serrano11@alumni.imperial.ac.uk
Abstract. 5G promises much faster Internet transmission rates at minimum la-
tencies with indoor and outdoor coverage; 5G potentially could replace tradi-
tional Wi-Fi for network connectivity and Bluetooth technology for geolocation
with a seamless radio coverage and network backbone that will accelerate new
services such as the Internet of Things (IoT). New infrastructure applications
will depend on 5G as a mobile Internet service provider therefore eliminating
the need to deploy additional private network infrastructure or mobile networks
to connect devices; however, this will increase cybersecurity risks as radio net-
works and mobile access channels will be shared between independent services.
To address this issue, this paper presents a digital channel authentication meth-
od based on the Blockchain Random Neural Network to increase Cybersecurity
against rogue 5G nodes; in addition, the proposed solution is applied to Physi-
cal Infrastructure: an Intelligent Building. The validation results demonstrate
that the addition of the Blockchain Neural Network provides a cybersecure
channel access control algorithm that identifies 5G rogue nodes where 5G node
identities are kept cryptographic and decentralized.
Keywords: Neural Network, Intelligent Buildings, Blockchain, Cybersecurity,
5G, Access Credentials ; Smart Infrastructure
1 Introduction
5G will be an innovative mobile solution that consists of very high carrier frequencies
with great bandwidth (mmWave), extreme node and device densities (ultra-
densification), and unprecedented numbers of MIMO antennas (massive multiple-
input multiple-output) [1]; potential new 5G application are divided into Enhanced
Mobile Broadband, Ultra Reliable Low Latency Communications, and Massive Ma-
chine Type Communications. 5G will be also highly integrative: the connection to a
5G air interface and spectrum together with LTE and WiFi will enable global, relia-
ble, scalable, available and cost-efficient connectivity solution that provides high rate
coverage and a seamless user experience; both are considered as a potentially key
driver for the yet to stablish global IoT [2]. To achieve these additional requirements,
the 5G core network will also have to reach extraordinary levels of flexibility and
intelligence to increase node capacity that supports high data rates, extremely low
latency, and a significant improvement in users’ perceived Quality of Service (QoS)
where energy and cost efficiencies will become even more critical considerations [3].
This high levels of network integration are provided at a cybersecurity cost, 5G
rogue nodes could gather user IoT data leaving users or devices incapable to use a
different transmission network such as Wi-Fi. Recent Blockchain solutions enable the
digitalization of contracts as it provides authentication between parties and infor-
mation encryption of data that gradually increments while it is processed in a decen-
tralized network such as the IoT [4]. Due to these features, Blockchain has been al-
ready applied in Cryptocurrency [5], Smart Contracts [6], Intelligent Transport Sys-
tems [7] and Smart Cities [8].
1.1 Research Motivation
To address the increased cybersecurity risk of 5G, this paper proposes a digital chan-
nel authentication method based on the Blockchain Random Neural Network [9,10].
The Blockchain Neural Network connects neurons in a chain configuration that pro-
vides an additional layer of resilience against Cybersecurity rogue 5G nodes in the
IoT. The 5G Cybersecurity application presented on this paper can be generalized to
emulate an Authentication, Authorization and Accounting (AAA) server where 5G
node identity is encrypted in the neural weights and stored decentralized servers.
The Blockchain Neural Network solution is equivalent to the Blockchain with the
same properties: 5G node identity authentication, data encryption and decentralization
where 5G node identities are gradually incremented and learned while are becoming
operational within the Intelligent Building. The Neural Network configuration have
analogue biological properties as the Blockchain where neurons are gradually incre-
mented and chained through synapses as variable 5G node identities increase; in addi-
tion, information is stored and codified in decentralized neural networks weights. The
main advantage of this research proposal is the biological simplicity of the solution
however it suffers high computational cost when neurons increase.
1.2 Research Proposal
5G and the Blockchain related work is described in Section 2. This paper defines a
decentralized solution that emulates the Blockchain validation process: mining the
input neurons until the neural network solution is found as presented in Section 3. The
proposed 5G node identity authentication mathematical model is described in Section
4; 5G nodes are authenticated before they are enabled into the Intelligent Building
where the International Mobile Subscriber Identity (IMSI) associated to the Intelligent
Building provides the Private Key and there is no need for a Public Key. Experi-
mental results in Section 5 show that the additional Blockchain neural network in-
creases cybersecurity resilience and decentralized confidentiality to 5G node identity
authentication. The main conclusion presented in Section 6 proves that the 5G node
identify is kept secret codified in the neural weights while detecting rogue 5G nodes.
2 Related work
2.1 Cybersecurity in 5G
5G will support a wide range of industry diverse business models and use cases ad-
dressing the motivations from different stakeholders such as mobile network opera-
tors, infrastructure and cloud service providers and tenants as presented by I. Adam
et al [11]; these diverse devices and applications will lead to different cybersecurity
requirements that shall be considered, in addition to performance requirements from
future applications as well as security and regulatory compliance with Service Level
Agreements in multi-tenant cloud environments. SHIELD is a novel design and de-
velopment cybersecurity framework which offers Security-as-a-Service in 5G as pre-
sented by D. Katsianis et al [12]; SHIELD framework leverages network functions
virtualization and Software Defined Networks for virtualization and dynamic place-
ment of virtual network security functions, Big Data analytics for real-time incident
detection and mitigation as well as trusted computing attestation techniques for secur-
ing both the infrastructure and its services.
Cyber physical systems create new online networking services and applications in
the IoT through their defined interactions and automated decisions as exposed by R.
Atat et al [13]; future 5G cellular networks will facilitate the physical systems com-
munications through different technologies such as device-to-device communications
that need to be protected from eavesdropping, especially with the large amount of
traffic that will constantly flow through the network. The connectivity of many stand-
alone IoT systems through the Internet or 5G Networks introduces numerous cyberse-
curity challenges as sensitive information is prone to be exposed to malicious users as
studied by B. Mozzaquatro et al [14]; appropriate security services adapted to the
threats improves IoT cybersecurity from an ontological analysis. The next wireless
strategy is centered on creative 5G and IoT, as confirmed by E. Chang et al [15];
while the confidence of the future mobile technology will help innovate governments,
workforce, industry and social media, the key threat to future mobile wireless devel-
opment is the growing concern of its security, privacy and anomaly detection. Fog
and mobile edge computing will play a key role in the upcoming 5G mobile networks
to support decentralized applications, data analytics and self network management by
using a highly distributed computing model, as stated by L. Fernández et al [16]; user-
centric cybersecurity solutions particularly require the collection, process and anal-
yses of significantly large amount of data traffic and huge number of network connec-
tions in 5G networks in real-time and in an autonomous way using Deep Learning
techniques to detect network anomalies as presented by L. Fernández et al [17].
The protection of the IoT is a challenging task due to system security is the founda-
tion for the development of IoT, as researched by Y. Lu et al [18]; the key factors of
the security model are the protection and integration of heterogeneous smart devices
and Information Communication Technologies (ICT). Mobile cloud computing is
applied in multiple industries to obtain cloud-based services by leveraging mobile
technologies. With the development of 5G, defensive mitigations against threats from
wireless communications have been playing a remarkable role in the Web security
domain and Intrusion Detection Systems (IDS) as shown by K. Gai et al [19]; a high
level framework for implementing secure mobile cloud computing that adopts IDS
techniques in 5G networks based on mobile cloud-based solutions.
2.2 Neural Networks and Cryptography
Neural Networks have been already applied to Cryptography; D. Pointcheval [20]
presents a linear scheme based on the Perceptron problem, or N-P problem, suited for
smart cards applications. W. Kinzel et al [21] train two multilayer neural networks on
their mutual output bits with discrete weights to achieve a synchronization that can be
applied to secret key exchange over a public channel; A. Klimov et al [22] propose
three cryptanalytic attacks (genetic, geometric and probabilistic) to the above neural
network. E. Volna et al [23] apply feed forward neural networks as an encryption and
decryption algorithm with a permanently changing key. A. Yayık et al [24] present a
two-stage cryptography multilayered neural network where the first stage generates
neural network-based pseudo random numbers and the second stage encrypts infor-
mation based on the non-linearity of the model. T. Schmidt et al [25] present a review
of the use of artificial neural networks in cryptography.
2.3 Blockchain and Security
Currently; there is a great research effort in blockchain algorithms applied to security
applications. D. Xu et al [26] propose a punishment scheme based on the action rec-
ord on the blockchain to suppress the attack motivation of the edge servers and the
mobile devices in the edge network. S.-C. Cha et al [27] utilize a blockchain network
as the underlying communication architecture to construct an ISO/IEC 15408-2 com-
pliant security auditing system. K. Gai et al [28] propose a conceptual model for fus-
ing blockchains and cloud computing over three deployment modes: Cloud over
Blockchain, Blockchain over Cloud and Mixed Blockchain-Cloud. Y. Gupta et al [29]
propose a Blockchain consensus model for implementing IoT security. R. Agrawal et
al [30] present a Blockchain mechanism that continuously evaluates legitimate pres-
ence of user in valid IoT-Zones without user intervention.
3 Blockchain Neural Network in 5G Cybersecurity
Blockchain [5] is based on cryptographic concepts which can be applied similarly by
the use of Neural Networks. Information in the Blockchain is contained in blocks that
also include a timestamp, the number of attempts to mine the block and the previous
block hash. Decentralized miners then calculate the hash of the current block to vali-
date it. Information contained in the Blockchain consists of transactions which are
authenticated by a signature that uses the Intelligent Building private key, transaction
origin, destination and value (Figure 1).
Hash
Time Stamp
Transactions
Iterations
Previous Hash
Hash
Time Stamp
Transactions
Iterations
Previous Hash
Hash
Time Stamp
Transactions
Iterations
Previous Hash
Block n-1 Block n Block n+1
Transaction
From
To
Data
Signature
Signature = Function (Private Key,
From, To, Value)
Verify Signature = Function (Public Key,
Signature, From, To, Value)
Fig. 1. Blockchain Model
3.1 The Random Neural Network
The proposed Blockchain configuration is based on the Random Neural Network
(RNN) [31-33] which is a spiking neuronal model that represents the signals transmit-
ted in biological neural networks, where they travel as spikes or impulses, rather than
as analogue signal levels. The RNN is a spiking recurrent stochastic model for neural
networks where its ma in analytica l properties are the “product form” and the exist-
ence of the unique network steady state solution.
3.2 The Random Neural Network with Blockchain configuration
The Random Neural Network with Blockchain configuration consists of L Input Neu-
rons, M hidden neurons and N output neurons Network (Figure 2). Information in this
model is contained networks weights w+(j,i) and w-(j,i) rather than neurons xL, zM, yN.
x1w+(j,i): excitatory
network weights
x2
xL
z1
z2
zM
y1
y2
yN
Input Layer Hidden Layer Output Layer
w-(j,i): inhibitory
network weights
Λ1
λ1
Λ2
λ2
ΛL
λL
Λ: External
excitatory signal
λ: External
inhibitory signal
External signals
i1
i2
iL
Fig. 2. The Random Neural Network structure
• I = (ΛL, λL), a variable L-dim ensional input vector I Є [-1,1]L represents the pair of
excitatory and inhibitory signals entering each input neuron respectively; where
scalar L values range 1<L< ∞;
• X = (x1, x2, … , xL), a variable L-dimensional vector X Є [0,1]L represents the
input state qL f or the neuron L; where scalar L values ra nge 1<L< ∞;
• Z = (z1, z2, … , zM), a M-dimensional vector Z Є [0,1]M that represents the hidden
neuron state qM for the neuron M; where scalar M values ra nge 1<M< ∞;
• Y = (y1, y2, … , yN), a N-dimensional vector Y Є [0,1]N that represents the neuron
output state qN for the neuron N; where scalar N values ra nge 1<N< ∞;
• w+(j,i) is the (L+M+N) x (L+M+N) matrix of weights that represents from the
excitatory spike emission from neuron i to neuron j; where i Є [xL, zM, yN] and j Є
[xL, zM, yN];
• w-(j,i) is the (L+M+N) x (L+M+N) matrix of weights that represents from the in-
hibitory spike emission from neuron i to neuron j; where i Є [xL, zM, yN] and j Є
[xL, zM, yN].
The main concept of the Random Neural Network Blockchain configuration is that
the neuron vector sizes, L, M and N are variable instead of fixed. Neurons or blocks
are iteratively added where the value of the additional neurons consists on both the
value of the additional information and the value of previous neurons therefore form-
ing a neural chain. Information in this model is transmitted in the matrixes of network
weighs, w+(j,i) and w-(j,i) rather than in the neurons. The input layer X represents the
Intelligent Building’s incremental verification data; the hidden layer Z represents the
values of the chain and the output layer Y represents the Intelligent Building Private
Key.
4 5G Cybersecurity Neural Network Blockchain Model
The 5G Cybersecurity Neural Network Blockchain model described in this section is
based on the main concepts shown on Figure 3:
• Private key, yN;
• Authentication, A(t) and Verification, V;
• Neural Chain network and Mining;
• Decentralized information, w+(j,i) and w-(j,i).
Floor 1
v1
No Mining
Decentralized Network
w+(j,i) w-(j,i)
Floor 2
v2
Mines xL and zM
Authentication 1
A(1) Authentication 2
A(2) Authentication 3
A(3)
Floor 3
v3
Mines xL and zM
Floor t
vt
Mines xL and zM
Intelligent Building
IMSI yN
Authentication t
A(t)
Fig. 3. 5G Cybersecurity Neural Network Blockchain Model
4.1 Private key
The private key Y = (y1, y2, … , yN) consists on the Intelligent Building International
Mobile Subscriber Identity (IMSI). The private key is presented by the Intelligent
Building every time a 5G node becomes operational and its identity credential re-
quires authorization and verification from the Intelligent Building in order to enable
its digital channel (Table 1).
Table 1. IMSI Private Key yN
Identifier
Code
Digits
Bits
Name
Key
IMSI
MCC
3
12
Mobile Country Code
y3
MNC
3
12
Mobile Network Code
y2
MSIN
9
36
Mobile Subscription Identification Number
y1
4.2 Authentication and Verification
Let’s define Authentication and Verification a s:
• Authentication, A(t) = {A(1), A(2), … A(t)} as a variable vector where t is the 5G
node authentication stage number;
• Verification, V = {v1, v2, … vt} as a set of t I-vectors where vo = (eo1, eo2, … eoI)
and eo are the I different dimensions for o=1,2, ... t.
The first Authentication A(1) has associated an input state X = xI which corresponds
v1 representing the 5G node identity data. The output state Y = yN corresponds to the
Intelligent Building Private Key and the hidden layer Z = zM corresponds to the value
of the neural chain that will be inserted in the input layer for the next authentication.
The second Authentication A(2) has associated an input state X = xI which corre-
sponds to the 5G node identity data v1 for the first Authentication A(1), the chain, or
the value of the hidden layer zM and the additional 5G node identity data v2. The out-
put state Y = yN still corresponds the Intelligent Building Private Key and the hidden
layer Z = zM corresponds to the value of the neural chain for the next transaction. This
process iterates as more 5G node identity data is inserted in a staged enabling process.
The neural chain can be formed of the values of the entire hidden layer neurons, a
selection of neurons, or any other combination to avoid the reverse engineering of the
Intelligent Building identity from the stored neural weights.
4.3 Neural Chain Network and Mining
The first Authentication A(1) calculates the Random Neural Network neural weights
with an Ek < Y for the input data I = (ΛL, λL), and the Intelligent Building private key
Y = yN. The calculated network weights w+(j,i) and w-(j,i) are stored in the decentral-
ized network and retrieved in the mining process. After the first Authentication; the
5G node requires to be validated at each additional Authentication with the Intelligent
Building private key where the 5G node verification data is validated and verification
data vt from Authentication A(t) are added to the Intelligent Building.
Verification data is validated or mined by calculating the outputs of the Random
Neural Network using the transmitted network weighs, w+(j,i) and w-(j,i) at variable
random inputs i, or following any other method. The solution is found or mined when
quadratic error function Ek is lesser than determined minimum error or threshold T.
Ek=1
2∑(y'n-yn)2
N
n=1
<T
(1)
where Ek is the minimum error or threshold, y’ is the output of the Random Neura l
Network with mining or random input I and yn is the Intelligent Building Private Key.
The mining complexity can be tuned by adjusting Ek. The Random Neural Network
with Blockchain configuration is mined when an Input I is found that delivers an out-
put Y with an error Ek lesser than a threshold T for the retrieved Intelligent Building
network weights w+(j,i) and w-(j,i).
When the solution is found, the Intelligent Building data can be processed, the po-
tential value of the neural hidden layer Z = zM is added to form the Neural Chain
where more Intelligent Building data is added. Once the solution is found or mined,
the values of the hidden layer are used in the input of the next transaction, along with
the new data; where the Random Neural Network with gradient descent learning algo-
rithm is calculated again to generate the new network matrixes w+(j,i) and w-(j,i). The
more authentication and verification data; the validation or mining process increases
on complexity.
Finally, the system calculates the Random Neural Network with Gradient descent
algorithm for the new pair (I, Y) where the new calculated network weights w+(j,i)
and w-(j,i) are stored in the decentralized network.
4.4 Decentralized Information
The Intelligent Building network weights w+(j,i) and w-(j,i) are stored in the decen-
tralized network rather than its data I directly from which are calculated. The network
weights expand as more verification data is inserted, therefore creating an adaptable
method. In addition; only the Intelligent Building Data can be extracted when the
IMSI private key is presented therefore making secure to store information in a decen-
tralized system.
5 Neural Blockchain in 5G Cybersecurity Validation
This section proposes a practical validation of the Neural Blockchain model in the 5G
Cybersecurity using the network simulator Omnet ++ with Java for a network of five
5G nodes with associated passive Distributed Antenna Systems. The experiment will
emulate a gradual implementation of a 5G network in an Intelligent Building were 5G
nodes are increasingly enabled into the Smart Building (Table 2).
Table 2. Neural Blockchain in Cybersecurity Validation – 5G Node values
Simulation
Application
Cell
Coverage
5G Node Identity
Vt
Intelligent Building
IMSI YN
Five
5G Nodes
Intelligent
Building
Pico
100 m
0.314km2
60 bits
234-151-234512340-4
60 bits
234-151-234512351
A 5G Network generic Performance Specification is shown on Table 3:
Table 3. 5G Network Performance Specification
Data Rate
Frequency
Latency
Modulation
Cell
20 Gbit/s
3.5 GHz - 26 GHz
<1 ms
Non-Orthogonal Multiple Access
Pico Small
The Intelligent Building is assigned to a IMSI private key yN whereas the gradual
addition of 5G nodes require validation before they are enabled to transmit infor-
mation into the Intelligent Building. When an 5G node is ready to become operation-
al, its identity is validated by the Intelligent Building; the decentralized system re-
trieves the neural weights associated to the private key; mines the block, adds the
node code and stores back the network weights in the decentralized system. Mining
for this validation is considered as the selection of random neuron values until Ek < T.
When more 5G nodes are ready to become operational, the Intelligent Building uses
its IMSI private key yN and the 5G node identity vt is added to the neural chain once it
is mined. Each bit is codified as a neuron however instead of the binary 0-1, neuron
potential is codified as 0.25 – 0.75 (Figure 4); this approach removes overfitting in the
learning algorithm as neurons only represent binary values.
V1 5G Node 1
0.25
0.75
1
0
bit Neuron Potential
1.0
0.5
0.0
Simulation Model – 5G Deployment Stages Bit - Neuron Codification
Intelligent
Building
IMSI yN V2 5G Node 2
V3 5G Node 3
V4 5G Node 4
V5 5G Node 5
A(1)
A(2)
A(3)
A(4)
A(5)
Fig. 4. Neural Blockchain in 5G Cybersecurity validation
The simulations are run 100 times for a five 5G node Network (Table 4). The infor-
mation shown is the number of iterations the Random Neuron Network with Block-
chain configuration requires to achieve an Ek < 1.0E-10; the error Ek, the number of
iterations to mine the Blockchain and the number of neurons for each layer; input xL,
hidden zM and output yN.
Table 4. 5G network Simulation – Learning and Mining
5G Node
Learning
Iteration
Learning
Error
Mining
Iteration
Mining
Threshold
Mining
Error Ek
Number of Neurons
(xL, zM, yN)
1
267.00
9.50E-11
106.87
1.00E-02
4.54E-03
60-4-60
2
219.73
9.48E-11
10230.04
1.00E-02
6.63E-03
124-4-60
3
194.76
9.38E-11
104456.1
1.00E-02
8.03E-03
188-4-60
4
178.91
9.29E-11
632011.8
1.00E-02
8.47E-03
252-4-60
5
167.22
9.45E-11
570045.3
1.00E-02
8.87E-03
316-4-60
With four neurons in the hidden layer, the number of learning iterations gradually
decreases while the number of input neurons increases due to the additional infor-
mation added activating 5G nodes. The results for the mining iteration are not as line-
ar as expected because mining is performed using random values (Figure 5). Surpris-
ingly; mining is easier in the final authentication stage when it would have been ex-
pected harder as the number of neurons increases.
Fig. 5. 5G Network Simulation – Learning and Mining Iterations
The Blockchain Random Neural Network algorithm must detect 5G rogue nodes to be
effective (Table 5); Δ represents the number of bit changes in the Node identity vt for
different values hidden neurons ZM.
Table 5. 5G Network Simulation – Rogue Node Tampering Error
5G
Node
Δ
Neurons
(xL, zM, yN)
Error
Neurons
(xL, zM, yN)
Error
Neurons
(xL, zM, yN)
Error
1
0.0
1.0
60-4-60
9.50E-11
1.83E-03
60-60-60
9.97E-11
1.53E-03
60-60-60
9.97E-11
1.53E-03
2
0.0
1.0
124-4-60
9.84E-11
2.43E-04
124-60-60
9.93E-11
1.40E-04
124-124-60
9.98E-11
1.40E-04
3
0.0
1.0
188-4-60
9.01E-11
6.76E-05
188-60-60
9.93E-11
3.57E-05
188-188-60
9.98E-11
3.78E-05
4
0.0
1.0
252-4-60
8.91E-11
2.58E-05
252-60-60
9.95E-11
1.30E-05
252-252-60
9.99E-11
1.48E-05
5
0.0
1.0
316-4-60
8.88E-11
1.57E-05
316-60-60
9.95E-11
6.08E-06
316-316-60
9.99E-11
4.54E-06
The addition of a 5G rogue node into the Intelligent Building is detected by the learn-
ing algorithm Neural Block Chain (Figure 6) even when the identity value only differ
in a bit, Δ=1.0.
ZM=4
ZM=YN
ZM=XL
ZM=XLZM=YNZM=4
Δ=0.0
ZM=XLZM=YNZM=4
Δ=1.0
Fig. 6. 5G Network Simulation – Rogue Node Tampering Error
The simulation results show that the increment of neurons in the hidden layer re-
quires additional learning iterations, however this increment is not reflected in higher
accuracy to detect input errors or 5G rogue nodes, therefore ZM=4 is the most opti-
mum configuration for this model.
6 Conclusions
This paper has presented the application of the Blockchain Random Neural Network
in 5G Cybersecurity where neurons are gradually incremented as 5G node identity
authentication data increases through their gradual addition into an Intelligent Build-
ing. This configuration provides the proposed algorithm the same properties as the
Blockchain: security and decentralization with the same validation process: mining
the input neurons until the neural network solution is found.
The Random Neural Network in Blockchain configuration has been applied to an
5G node Authentication server; experimental results show that Blockchain applica-
tions can be successfully implemented using neural networks where mining effort can
be gradually increased, user authentication and data encryption in a decentralized
network therefore removing centralized validation mechanisms.
This paper has proposed a Cybersecurity application in 5G where the addition of a
5G node to the Intelligent Building requires prior authentication and verification be-
tween decentralized parties. 5G node identity data is encrypted, information is decen-
tralized where a 5G rogue node attackers can be identified if it is present in the Intel-
ligent Building.
Future work will include the authentication and verification of additional 5G nodes
and the impact of the number of chain neurons in relation to the learning and mining
process. In addition, the mining threshold will be assessed against the required mining
iterations that find the network solution.
References
1. M. Palattella, M. Dohler, A. Grieco, G. Rizzo, J. Torsner, T. Engel, L. Ladid. Internet of
Things in the 5G Era: Enablers, Architecture, and Business Models. IEEE Journal on Se-
lected Areas in Communications, 34, 3. (2016), 510 – 527.
2. J. Andrews, S. Buzzi, W. Choi, S. Hanly, A. Lozano, A. Soong, J. Zhang. What Will 5G
Be?. IEEE Journal on Selected Areas in Communications, 32, 6. (2014), 1065 – 1082.
3. M. Agiwal, A. Roy, N. Saxena. Next Generation 5G Wireless Networks: A Comprehen-
sive Survey. IEEE Communications Surveys & Tutorials, 18, 3. (2016), 1617 – 1655.
4. S. Huh, S. Cho, S. Kim. Managing IoT devices using Blockchain platform. International
Conference on Advanced Communication Technology. (2017), 464-467.
5. S Nakamoto. Bitcoin: A Peer-to-Peer Electronic Cash System - Bitcoin.org. (2008), 1-9.
6. H. Watanabe, S. Fujimura, A. Nakadaira, Y. Miyazaki, A. Akutsu, J. Kishigami. Block-
chain contract: Securing a blockchain applied to smart contracts. International Conference
on Consumer Electronics. (2016), 467 -468.
7. Y. Yuan, F.-Y. Wang. Towards blockchain-based intelligent transportation systems. Inter-
national Conference on Intelligent Transportation Systems. (2016), 2663-2668.
8. K. Biswas, V. Muthukkumarasamy. Securing Smart Cities Using Blockchain Technology.
International Conference High Performance Computing and Communications / Smart City
/ Data Science and Systems. (2016), 1392-1393.
9. W. Serrano. The Random Neural Network with a Blockchain Configuration in Digital
Documentation. International Symposium on Computer and Information Sciences. (2018),
196-210.
10. W. Serrano. The Blockchain Random Neural Network in Cybersecurity and the Internet of
Things. Artificial Intelligence Applications and Innovations. (2019), 50-63.
11. I. Adam, J. Ping. Framework for Security Event Management in 5G. International Confer-
ence on Availability, Reliability and Security, 51. (2018), 1-10.
12. D. Katsianis, I. Neokosmidis, A. Pastor, L. Jacquin, G. Gardikis. Factors Influencing Mar-
ket Adoption and Evolution of NFV/SDN Cybersecurity Solutions. Evidence from
SHIELD Project. European Conference on Networks and Communications. (2018), 261-
265
13. R. Atat, L. Liu, H. Chen, J. Wu, H. Li, Y. Yi. Enabling cyber-physical communication in
5G cellular networks: challenges, spatial spectrum sensing, and cyber-security. IET Cyber-
Physical Systems: Theory & Applications, 2, 1. (2017), 49 – 54.
14. B. Mozzaquatro, C. Agostinho, D. Goncalves, J. Martins, R. Jardim. An Ontology-Based
Cybersecurity Framework for the Internet of Things. Sensors, 18, 9, 3053. (2018), 1-20.
15. E. Chang, F. Gottwalt, Y. Zhang. Cyber Situational Awareness for CPS, 5G and IoT. Fron-
tiers in Electronic Technologies, 433. (2017), 147-161.
16. L. Fernández, A. Huertas, C. Manuel Gil, F. Garcia, G. Martinez. Dynamic management of
a deep learning-based anomaly detection system for 5G networks. Journal of Ambient In-
telligence and Humanized Computing. (2018), 1–15.
17. L. Fernandez, A. Perales, F. Garcia, M. Gil, G. Martinez. A Self-Adaptive Deep Learning-
Based System for Anomaly Detection in 5G Networks IEEE Access, 6. (2018), 7700 –
7712.
18. Y. Lu, L. Da-Xu. Internet of Things (IoT) Cybersecurity Research: A Review of Current
Research Topics. IEEE Internet of Things Journal. (2018), 1-13.
19. K. Gai, M. Qiu, L. Tao, Y. Zhu. Intrusion detection techniques for mobile cloud compu-
ting in heterogeneous 5G. Security Comm. Networks, 9. (2016), 3049-3058.
20. D. Pointcheval. Neural Networks and their Cryptographic Applications. Livre d es resumes
Eurocode Institute for Research in Computer Science and Automation. (1994), 1-7.
21. W. Kinzel, I. Kanter. Interacting Neural Networks and Cryptography Secure exchange of
information by synchronization of neural networks. Advances in Solid State Phy sic, 42.
(2002), 383-391.
22. A. Klimov, A. Mityagin, A. Shamir. Analysis of Neural Cryptography. International Con-
ference on the Theory and Application of Cryptology and Information Security, 2501.
(2002), 288-298.
23. E. Volna, M. Kotyrba, V. Kocian, M. Janosek. Cryptography based on the Neural Net-
work. European Conference on Modelling and Simulation. (2012), 1-6.
24. A. Yayık, Y. Kutlu. Neural Network based cryptography. International Journal on Neural
and Mass - Parallel Computing and Information Systems, 24, 2. (2014), 177-192.
25. T. Schmidt, H. Rahnama, A. Sadeghian. A review of applications of artificial neural net-
works in cryptosystems. World Automation Congress. (2008), 1 -6.
26. D. Xu, L. Xiao, L. Sun, M. Lei. Game theoretic study on blockchain based secure edge
networks. IEEE International Conference on Communications in China. (2017), 1 – 5.
27. S.-C. Cha, K.-H. Yeh. An ISO/IEC 15408-2 Compliant Security Auditing System with
Blockchain Technology. IEEE Conference on Communications and Network Security.
(2018), 1-2.
28. K. Gai, K.-K. Raymond, L. Zhu. Blockchain-Enabled Reengineering of Cloud Datacen-
ters. IEEE Cloud Computing, 5, 6. (2018), 21-25.
29. Y. Gupta, R. Shorey, D. Kulkarni, J. Tew. The applicability of blockchain in the Internet
of Things. IEEE International Conference on Communication Systems & Networks.
(2018), 561 – 564.
30. R. Agrawal, P. Verma, R. Sonanis, U. Goel, A. De, S. Anirudh, S. Shekhar. Continuous
Security in IoT Using Blockchain. IEEE International Conference on Acoustics, Speech
and Signal Processing. (2018), 6423 – 6427.
31. E. Gelenbe. Random Neural Networks with Negative and Positive Signals and Product
Form Solution. Neural Computation, 1. (1989), 502-510.
32. E. Gelenbe. Learning in the Recurrent Random Neural Network. Neural Computation, 5.
(1993), 154-164.
33. E. Gelenbe. G-Networks with Triggered Customer Movement. Journal of Applied Proba-
bility, 30. (1993), 742-748.
Appendix
x1
x2
z1
z2
x3
x4
z3
z4
x5
x6
z5
z6
x1
x2
z1
z2
IL HL OL
Authentication 1 A(1) Authentication 2 A(2)
IL HL OL
x1
x2
z1
z2
x3
x4
z3
z4
x5
x6
z5
z6
Authentication 3 A(3)
x7
x8
z7
z8
x9
x10
z9
z10
x11
x12
z11
z12
x13
x14
z13
z14
IL HL OL
Chain Neuron
Data Neuron
x1,x2 = Data Roaming 1
z1,z2 = Chain Roaming 1
y1,y2 = Private Key
x1,x2 Data Roaming 1
x3,x4 Chain Roaming 1-2
x5,x6 Data Roaming 2
z1-z6 = Chain Roaming 2
y1,y2 = Private Key
Legend
IL: Input Layer
HL: Hidden Layer
OL: Output Layer
x1,x2 Data Transaction 1
x3,x4 Chain Transaction 1-2
x5,x6 Data Transaction 2
z1-z6 = Chain Transaction 2
x7-x12 Chain Transaction 2-3
x13,x14 Data Transaction 3
y1,y2 = Private Key
Key Neuron
y1
y2
y1
y2
y1
y2
Mining the Network:
Find i1 ... iL that make Ek<T
Verification
Authentication 1
v1
Verification Authentication 2
v2
Verification Authentication 3
v3
