Cloud and MEC Security 389
16.5.1 Denial of Service (DoS) Attack
DoS attack is an age‐long threat in various computing and networking arenas. DoS
creates an artificial scarcity or lack of online resources and network services. DoS
attacks could happen in the form of distributed denial‐of‐service (DDoS) or wireless
jamming and could be launched on both the virtualization and network infrastructures.
In the case of MEC systems, DoS attacks have limited scope, as described in [38], as a
DoS attack on the network edge will affect only the attacked vicinity and not the entire
network. So also an attack on the core network infrastructure might not lead to a
complete disruption in the functionality of the edge data centers. This is due to the
autonomous and semi‐autonomous nature of their protocols and services.
Another loophole in the MEC architecture that makes it vulnerable to DoS attack is the
combination of multiple Virtual Machines (VMs) spread across several mobile edge hosts.
This formation increases the possibilities of compromising multiple VMs simultaneously,
leading to large‐scale attacks such as DDoS [39]. On the pros side, the MEC mobile network
infrastructure by virtue of its design is inherently suited for deploying extended defense
perimeters capable of mitigating the DDoS attack within multiple fronts. Such defense
mechanism allows for fragmented deployments at the network edge, hence capable of
defending against smaller attacks before they get any further chances of escalating.
16.5.2 Man‐in‐the‐Middle (MitM)
The MitM attack is characterized by the presence of a third malicious party interposed
between two or more communicating parties or entities and secretly relaying or altering
the communication between such parties; a common example is the MitM attack
between a server and a client. For the MEC scenario, a MitM attack is categorized as an
infrastructure attack [38], where the malicious attacker tries to hijack a certain segment
of the network and begins to launch attacks, such as eavesdropping and phishing, on
connected devices.
The potency of an MitM attack on mobile networks has been proven in various works
and literature [40,41]. In these works, MitM attack was launched between 3G and
WLAN networks. Such attacks would be even more threatening for the MEC scenario,
given that MEC relies heavily on virtualization, hence launching an MitM attack on
multiple VMs could very easily affect all other elements on both sides of the attack.
16.5.3 Inconsistent Security Policies
In mobile networks, the need to preserve user‐security parameters when they roam
from one operator network to another is of utmost importance. In the case of MEC, it
is highly possible that all the security services are not updated very frequently and per‐
user basis. MEC servers can also be limited with resources, thus making it more challenging to facilitate such services. When a user moves from one operator network to
another, latency sensitive services are utilized, and such user might be provided with
these services through the visited operator MEC, but will his security or the security of
the service he is using be ensured? This buttresses the need for security policy sharing
among network operators on a much faster scale, to ensure that users traffic on the
access networks are effectively attached to the MEC their services are migrating to.
16.5.1 Denial of Service (DoS) Attack
DoS attack is an age‐long threat in various computing and networking arenas. DoS
creates an artificial scarcity or lack of online resources and network services. DoS
attacks could happen in the form of distributed denial‐of‐service (DDoS) or wireless
jamming and could be launched on both the virtualization and network infrastructures.
In the case of MEC systems, DoS attacks have limited scope, as described in [38], as a
DoS attack on the network edge will affect only the attacked vicinity and not the entire
network. So also an attack on the core network infrastructure might not lead to a
complete disruption in the functionality of the edge data centers. This is due to the
autonomous and semi‐autonomous nature of their protocols and services.
Another loophole in the MEC architecture that makes it vulnerable to DoS attack is the
combination of multiple Virtual Machines (VMs) spread across several mobile edge hosts.
This formation increases the possibilities of compromising multiple VMs simultaneously,
leading to large‐scale attacks such as DDoS [39]. On the pros side, the MEC mobile network
infrastructure by virtue of its design is inherently suited for deploying extended defense
perimeters capable of mitigating the DDoS attack within multiple fronts. Such defense
mechanism allows for fragmented deployments at the network edge, hence capable of
defending against smaller attacks before they get any further chances of escalating.
16.5.2 Man‐in‐the‐Middle (MitM)
The MitM attack is characterized by the presence of a third malicious party interposed
between two or more communicating parties or entities and secretly relaying or altering
the communication between such parties; a common example is the MitM attack
between a server and a client. For the MEC scenario, a MitM attack is categorized as an
infrastructure attack [38], where the malicious attacker tries to hijack a certain segment
of the network and begins to launch attacks, such as eavesdropping and phishing, on
connected devices.
The potency of an MitM attack on mobile networks has been proven in various works
and literature [40,41]. In these works, MitM attack was launched between 3G and
WLAN networks. Such attacks would be even more threatening for the MEC scenario,
given that MEC relies heavily on virtualization, hence launching an MitM attack on
multiple VMs could very easily affect all other elements on both sides of the attack.
16.5.3 Inconsistent Security Policies
In mobile networks, the need to preserve user‐security parameters when they roam
from one operator network to another is of utmost importance. In the case of MEC, it
is highly possible that all the security services are not updated very frequently and per‐
user basis. MEC servers can also be limited with resources, thus making it more challenging to facilitate such services. When a user moves from one operator network to
another, latency sensitive services are utilized, and such user might be provided with
these services through the visited operator MEC, but will his security or the security of
the service he is using be ensured? This buttresses the need for security policy sharing
among network operators on a much faster scale, to ensure that users traffic on the
access networks are effectively attached to the MEC their services are migrating to.