Security Challenges in Cyber Physical Systems and Internet of Things Systems
Security Challenges in Cyber Physical Systems and Internet of Things Systems
Security Challenges in Cyber Physical Systems and Internet of Things Systems
of things systems
Introduction:
Safety and security are the major challenges for cyber physical systems because of the serious
consequences and large attack surface. CPSs operate on physical plants which damages
themselves, people, or property when they are improperly used . .Security which is depended on
privacy is are concerned on the information. Recently it is considered as privacy is the property
of safety where security is its major need. False data entry is also constitute to an attack.
Computing system provides many models for separating the data .in addition to them
some threat models to be created in order to show the vulnerabilities. CPSs and IoT combine
physical plants and computational process and this coupling is known as the application
programming interfaces(APIs). These provide bidirectional communication between the cyber
and physical plants. This coupling is worrisome because it increases the fault rate in both the
cyber and physical plants. Many computational methods have been introduced for analyzing
systems and to identify vulnerabilities and enable resilient system design.
Environmental Tiers: consists of physical devices and a target environment which includes
end-users using the devices and their associated physical environment.
Service Tiers: a typical computing environment with services in SOA and CC (Cloud
Computing).
Control Tiers: to receive monitored data which are gathered though sensors, to make controlling
decisions, to find right services by consulting service framework, and to let the services invoked
on the physical device. The architecture is in Figure 1.
Cyber world and physical world are different essentially, but they are connected and affect each
other by information. One of the main features of physical world is dynamic, the same entity at
different time showed different properties. Therefore, when modeling the physical world entities,
the dynamic features should be considered in particular. In the cyber world, changes are
represented by state transitions, thus, simulating the physical world may lead to state explosion.
This is an important feature to be considered in the modeling and design process of CPS
fig2: CPS architecture based on SOA
Computing systems provide several models for sharing and separation of data . Distributed
computing provides models for shared and separate memory spaces. Fault models for
computing systems . In addition, a set of threat models and attack surfaces can be created to
describe the systems’ vulnerability to malicious activity.
CPSs and IoT systems couple physical plants and computational processes. That coupling
may be defined as models such as application programming interfaces (APIs). APIs can, in
general, provide bidirectional communication between the physical and cyber plants.
This coupling is worrisome in itself because it enlarges the fault and threat models for both the
cyber and physical subsystems. Although most methods have been focusing on network
topological characteristics and statistical analyses of data, following traditional approaches to
attack and failure detection, recently, computational formal methods have been introduced to
analyze systems, identify vulnerabilities, and enable resilient system design.
CPS ARCHITECTURE BASED ON SOA:
SOA(service-oriented architecture) is a dynamically integrating paradigm, which is used for
integrating loosely coupled services into one workflow using a standards-based software
component technology. Based on the SOA architecture, we proposed the generic CPS
architecture. The architecture is described as the following figure 2
IOT-ARCHITECTURE:
A.)The Basic Model :
The most prominent model known for IoT is the 3- layer architecture as shown in Fig. 3
consisting of the Perception Layer, the Network Layer and the Application Layer
1. Perception Layer-This layer is also known as the Device layer as it consists of
objects/devices and sensors. It acts as the skin of a living organism and its five senses to
identify objects, collect data and information. It includes technologies like RFID tags and
readers/writers, 2-D bar code labels, GPS,sensor networks and terminals.
2.Network Layer-This layer is also called the Transmission Layer, used to transmit and
process data provided by the Perception Layer. It works as the neural network and brain of a
human body. Technologies that work in this layer are 3G, WiFi, Bluetooth, Infrared, Zigbee
etc. It helps in management of network, information service center and network center
3.Application Layer -This layer emphasises on the industrial need and social division of IoT.
The main motive is to realize the level of intellectualization of industry. It covers various
virtual markets.
Fig 3:Basic model of IoT
5 layers are:
1)Object Layer
2) Object Abstraction Layer
3) Service Management Layer
4) Application Layer
5) Business Layer
FIG 5:Buissiness layer of IoT
Differences of IoT and cyber physical systems:
IoT is something which is closely associated with the fact that how can physical devices be
connected with each other over the internet to do something useful. E.g. The automated water
supply using the smartphone app and so on. So the major aspect for IoT becomes the physical
devices and the Internet. Generally IoT is kept limited to control & monitor small devices. Such
systems have no constraints towards having the devices on a common network.
On the other hand cyber physical systems are more concerned as to how physical systems be
monitored and controlled using the cyber-space. And such physical system are generally
experimental physical systems. E.g. Atomic reactors are great example of cyber physical
systems. Such system generally need the devices and the systems on a common network. How
such systems are not forced to follow such constraint, but the industry and the community
follows them based on the requirements for the system. Some very good examples for such
system could be Radio telescopes, Thermo-nuclear reactors, Power station monitoring & control
systems etc.
At a higher level view both these domain show remarkable similarity in their solution
architecture. But these domains are different in terms of their implementation and usage.
Challenges:
The physical plants of CPS and IoT applications vary widely, they all make common use of a
core set of computational components: processors, networks, operating systems, and middleware.
Making these systems safe and secure is especially difficult because they are long-lived systems
that cannot be easily replaced or, in some cases, even shut down Moore’s law and constantly
improving semiconductor technology have led to quick upgrade cycles that allow computer
scientists to avoid many of the problems associated with longevity. A variety of guidelines and
standards have been developed for the design of reliable CPSs. The MISRA C Guidelines for
automotive software. These coding guidelines are divided into mandatory, required and advisory
levels.
A.Real-time control
Real-time data processing: process queries within certain deadlines or response time bounds
.
Reliable event detection: support highly accurate event detection despite potentially noisy,
faulty, or compromised sensor data.
Real-time routing: sensor data must be delivered to nRTEDBs in a timely fashion. Also,
nRTEDBs need to efficiently exchange information with each other, while controlling sensors.
Security and robustness: to avoid the single point of attack problem discussed before in
addition to data confidentiality, integrith, and authenticity.
B. Security Assurance:
Confidentiality, integrity and availability are three important problems of system security that
need to be considered.We divide CPS into two categories, which are security critical system and
non-security critical system. Then to the former, confidentiality, integrity and availability are
very significant, and at the same time, other requirements such as physical security are also
emphasized. But to the later, integrity and availability are underlined and security is relatively
weakened. There are different requirements in different scenarios. For example, as for military
applications, the confidentiality feature is important, but in the smart home system, the real-time
requirements are emphasized.Security of CPS can be divided into the following three aspects:
perception security, which is to ensure the security and accuracy of the information collected
from physical environment; transport security, which is to prevent the data from being destroyed
during the transmission processes; processing center security, such as physical security and
safety procedures in servers or workstations.
Networking: deal with the data aggregation, diffusion. Various sensors could generate much
data which is to be aggregated or diffusion for analyzers to process further.
Computing: this step is for reasoning and analyzing the data collected during monitoring to
check whether the physical process satisfies certain pre-defined criteria.
Actuation: execute the actions determined during the computing phase.
Based on this, a context-aware security framework for CPS was proposed, as shown in Figure 3
Fig 3:Context-aware security framework for CPS.
C. Integration mechanism
The integration of CPS means coordination of different terminal sensors, mobile devices,
computer workstations, and networks. CPS integration has four features:
Design time techniques:Many design time and runtime techniques have been developed in
order to ensure the safety and security I CPSs and IoT.
1. Model based designing which emerged as the main stream technique in CPSs design.
The combination of formal modeling and simulation supported by model-based design
allows for more thorough checks of safety and security properties. 2. Architectures for
CPSs can be designed to ensure that properties related to safety and
security are provided by the system and respected by its components.
3. System Synthesis Algorithms implementations of CPSs can be used to ensure that
specifications and architectures are correctly translated, avoiding the introduction of errors
that can compromise safety or security. 4. Hardware Trojans a circuit that implements an
undesirable or malicious function. The
hardware Trojan is inserted into a larger circuit in order to alter its function.
Runtime techniques:
Finally the conclusion is that the systems that combine both physical and computer components
are becoming more complex. This systems should satisfy both safety and security properties.
But in many cases the systems fall short of our expectations and results in failures. Improving
the safety and security of CPSs and IoT systems requires the application of new techniques at
both design time and runtime as well as the diligent application of existing best practices. We
don’t have single solution to solve all these safety and security problems at once but we have
many methods in order to have the safety and security of out systems.
References:
D. N. Serpanos and A. G. Voyiatzis, “Security challenges in embedded systems,” ACM Trans. Embedded
Comput. Syst., vol. 12, no. 1s, Mar. 2013 C. Koulamas, S. Giannoulis, and A. Fournaris, “IoT
components for secure smart building environments,” in Components and Services for IoT Platforms:
Paving the Way for IoT Standards. New York, NY, USA: Springer-Verlag, 2016 H. Yoo and T. Shon,
“Grammar-based adaptive fuzzing: Evaluation on SCADA modbus protocol,” in Proc. IEEE Int. Conf.
Smart Grid Commun., Nov. 2016, pp. 557–563. M. U. Tariq, J. Florence, and M. Wolf, “Improving the
safety and security of wide-area cyber-physical systems through a resource-aware, service-oriented
development methodology,” in Proc. IEEE, vol. 106, no. 1, Jan. 2018, DOI: 10.1109/
JPROC.2017.2744645.
B. Cui, S. Liang, S. Chen, B. Zhao, and X. Liang, “A novel fuzzing method for Zigbee based on finite
state machine,” Int. J. Distrib. Sensor Netw., vol. 10, no. 1, p. 762891, 2014.
Information Technology—Message Queuing Telemetry Transport (MQTT) v3.1.1, document ISO/IEC
20922:2016.