US 12,010,518 B2 · System and method for securely defending against collusive attack under internet of vehicles

Abstract

A system and method for securely defensing against a collusion attack under Internet of Vehicles (IoV) are provided. The present disclosure can repair a vulnerability, of a reputation system in the IoV, that the IOC attackers can manipulate a traffic-related message aggregation model (TMAM) by increasing their own reputation scores in an inside-and-outside collusion (IOC) manner. In addition, the present disclosure can detect IOC attacks quickly to improve the security of the IoV; can eliminate suspicious providers recursively and provide a reputation fluctuation association rule, to avoid overload of the TMAM; and can deprive IOC attackers of the opportunity to improve their reputation scores and ensure credible information in the IoV, to ensure the fairness and availability of the TMAM without the interference from the IOC attackers.

A method for securely defensing against a collusion attack under Internet of Vehicles (IoV), performed on a cloud server of an IoV platform or a roadside unit (RSU) in a distributed IoV, and comprising:

detecting an inside-and-outside collusion (IOC) attack by using a reputation fluctuation association analysis (RFAA) method, analyzing a reputation fluctuation characteristic, and deleting an unlikely IOC provider from all providers in a current event voting operation consuming traffic,
wherein the RFAA comprises:
(1) supporting data management for detecting the IOC attack, wherein the supporting data management for detecting the IOC attack comprises two-step-based processing message report and consumer rating, and methods of the two-step-based processing message report and consumer rating are specifically as follows:
a TMAM stores a historical message from a provider and historical rating data of a consumer at the end of a voting operation consuming traffic, the provider provides a message for an on-board unit of a vehicle, and the consumer receives a message for the on-board unit of the vehicle and rates the message to complete the two-step-based processing message report;
and
each vehicle Vi is assigned to a P-C table, the P-C table stores the message reported previously by the provider on the vehicle, messages provided by the all providers, and ratings of Vi when the vehicle serves as the consumer, to complete the consumer rating, and
(2) obtaining a reputation fluctuation association rule, wherein the obtaining reputation fluctuation association rule supports the detection of the IOC attack, and the reputation fluctuation association rule is obtained by analyzing a characteristic of the IOC attack;
wherein the obtaining a reputation fluctuation association rule comprises:
step 1-1:
initializing a value of an index variable μi of a reputation fluctuation index to 0;
step 1-2:
setting an initial value of an integer variable k to 1;
step 1-3:
comparing the value of the integer variable k with h, wherein if k>h, the comparison is finished, and the method of obtaining the reputation fluctuation association rule is obtained;
and if k≤h, Rki≥ε+η2, Rki≤Rk−1i, Rki continuously decreases, and Rki≤ε+η1,μi=μi+1, wherein h is a quantity of voting actions that are related to a traffic event and initiated by a vehicle Vi, Rki represents a kth reputation score of the vehicle Vi, ε as a reputation score threshold is always set to a moderate value in a range of 0 to 1, a value range of η1 as a reputation warning line is 0≤η1<ε, and a value range of η2 as a high reputation line is η1<η2≤ε.
step 1-4:
if Rki≤ε+η1, Rki≥Rk−1i, Rki continuously decreases, and Rki≥ε+η2, μi=μi+1;
and
step 1-5:
returning to step 1-3 if k=k+1;
wherein the detecting an IOC attack by using a RFAA method further comprises analyzing an association relationship between IOC attackers, and a method for analyzing the association relationship between the IOC attackers is specifically as follows:
IOC providers are frequent providers who always appear together, and when the association relationship between the IOC attackers is analyzed, an index support numbers (Ø) is used to identify the frequent providers;
an association relationship between an IOC consumer and an IOC provider is described, it can be found that an IOC user and the IOC provider always appear at the same time, and the IOC consumer who always appears with the IOC provider at the same time are determined;
and
if a quantity of the IOC providers is more than half of a quantity of providers in the current voting operation of a traffic-related event, the IOC attackers can successfully increase their reputation scores,
if Vi is a consumer, and Φ is a group of providers in the current event voting operation consuming traffic,
Φ
1
2
is used to represent a set of the IOC providers, a quantity of
Φ
1
2
elements is at least half of a quantity of elements according to a principle of majority decision, the quantity of
Φ
1
2
elements is set as in formula (4):
❘
"\[LeftBracketingBar]"
Φ
1
2
❘
"\[RightBracketingBar]"
=
⌊
❘
"\[LeftBracketingBar]"
Φ
❘
"\[RightBracketingBar]"
2
⌋
+
1
,
(
4
)
for each
V
j
∈
Φ
1
2
,
the reputation fluctuation association rule is designed as shown in formula (5):
s
⁢
(
V
i
→
Φ
1
2
)
≥
minsup
,
and
(
5
)
formula (5) is under a constraint of μi≥2, wherein μi represents an index variable of a reputation fluctuation index of the provider Vi, Vi is the consumer, Vi is the provider, and minsup is a minimum value of support counts;
and a provider meets formula (5) is classified as an attacker;
wherein the method for analyzing the association relationship between the IOC attackers further comprises a dynamic sampling observation method, comprising:
step 2-1:
setting an initial value of minsup to a preset value;
step 2-2:
checking a quantity h of detected IOC attacks;
step 2-3:
outputting an average value of support counts by using a
mean
1
≤
h
≤
m
(
s
⁡
(
V
i
→
Φ
1
2
)
)
h
function, obtaining a minimum support count by using a
min
1
≤
h
≤
m
(
s
⁡
(
V
i
→
Φ
1
2
)
)
h
function, updating the value of minsup by using the minimum support count, and proceeding to step 2-5;
step 2-4:
determining, by using
mean
1
≤
h
≤
m
(
s
⁡
(
V
i
→
Φ
1
2
)
)
h
≤
mean
1
≤
h
≤
m
⁢
(
s
⁡
(
V
i
→
Φ
1
2
)
)
h
′
,
whether a determining formula is valid, performing step 2-5 if the determining formula is invalid, obtaining the minimum support count by using a
min
1
≤
h
≤
m
⁢
(
s
⁡
(
V
i
→
Φ
1
2
)
)
h
′
function if the determining formula is valid, updating the value of minsup by using the minimum support count, and proceeding to step 2-5;
and
step 2-5:
checking another h detected IOC attacks, obtaining an average value of support counts by using a
mean
1
≤
h
≤
m
⁢
(
s
⁡
(
V
i
→
Φ
1
2
)
)
h
′
function, and returning to step 2-4, wherein i, h, and m are natural numbers.

224 more elements

System and method for securely defending against collusive attack under internet of vehicles

Life of this patent

System and method for securely defending against collusive attack under internet of vehicles

Life of this patent

Abstract

Claims3