Page 60 - 卫星导航2021年第1-2合期

Chen et al. Satell Navig (2021) 2:12 Page 5 of 7

Optoelectronics have carried out the research on NMA Table 1 Performance comparison of diferent Schemes
authentication (Liu 2015, 2018; Mu et al. 2020). Te Schemes C/N Avg. TBA Max. TBA Max. AL
0
existing SBAS signal authentication protocol proposed in (AER = 1%) (AER 1%) (s) (s)
Europe and the United States are based on the traditional (dB·Hz) (s)
bent-pipe systems such as WAAS and EGNOS, while the L5-I ECDSA 28.5 13.52 18 20–29
Chinese BeiDou Satellite Based Augmentation System
(BDSBAS) adopts on-board signal generation system, L5-I TESLA 28.3 5.9 6 11
which is somewhat diferent from WAAS and EGNOS. L1-I ECDSA 28.5 12.89 18 20–29
In 2019, BDSBAS authentication message and simulation L1-I TESLA 28.3 4.89 6 11
experiments were carried out by Mu et al. of the China Q ECDSA I/Q 1:1 31.3 3.03 3 4
Academy of Sciences. Te message was designed based Q TESLA I/Q 1:1 31 1.01 1 1
on China’s commercial cryptography system standard Q ECDSA I/Q 3:1 29.1 5.05 5 8
SM2 (GM/T 32918-2016) and ECC algorithm (Standardi- Q TESLA I/Q 3:1 29.3 2.02 2 4
zation Administration, 2016a, 2016b), and the simulation
verifcation of Over Te Air Rekeying (OTAR) broadcast-
ing process was carried out. L1/L5 scenarios, were conducted in European air service
area (Fernández‐Hernández et al., 2018).
Table 2 summarizes the impact of TESLA and ECDSA
Simulation results and analysis schemes on the service performances such as Vertical
Te simulation trials based on the EGNOS EAST plat- Position Errors (VPE), Vertical Protection Level (VPL),
form were carried out by Fernandez-Hernandez et al. continuity, and availability of SBAS under diferent Page
(Fernandez-Hernandez et al. 2014), and the results on the Error Rate(PER) conditions. When PER = 0, the presence
performances of SBAS authentication in the I/Q-channel or absence of authentication has no efect on all perfor-
schemes, as well as SBAS authentication were presented.
–3
mance indicators. For PER = 1 × 10 , since the loss of
each message may cause identity authentication failure,
Simulation results of authentication performance the continuity risk of SBAS messages after joining the
in I/Q‑channel schemes authentication protocol is signifcantly higher, but the
According to the 128-bit security level, ECDSA authen- availability remains above 99%. It can be seen that join-
tication message (512-bit) requires three 216-bit mes- ing the authentication service will have an impact on the
sage frames, but TESLA only needs one 216-bit message SBAS message but still meet the availability performance.
frame. At this time, the maximum TBA of TESLA is
six seconds, and the maximum TBA of ECDSA is 18 s.
Table 1 shows the simulation results of the SBAS message Conclusion
authentication schemes. Tis article introduces two diferent SBAS message
For L1-ECDSA, a 1% Authentication Error Rate authentication methods, ECDSA and TESLA, and four
(AER) is achieved with a Carrier-to-Noise ratio (C/N ) diferent feasible schemes combined with the current
0
of 28.5 dB·Hz. In these conditions, the average TBA is SBAS Interface Control Document (ICD). Combined
13.52 s, the authentication period is 18 s (three message with the simulation results of European EGNOS in
frames), and the maximum AL is from 20 to 29 s. Te EAST, the results of several performance indicators with
maximum delay suggests that due to the 1% of authen- or without certifcation are analyzed. It can be seen that
tication failures, three digital signature message frames after joining the authentication service, the performance
may have an additional digital signature frame. of SBAS is less afected. SBAS messages are protected
Te 6s TTA required by SBAS is just satisfed in the against spoofng.
Q channel scheme. Using I/Q power 1:1 allocation will Starting from improving the design of signals, SBAS
reduce the performance. A power apportionment of authentication provides user terminals with the technical
75%/25% for the I/Q channels will reduce the Q chan- means to cope with spoofng and interference, enhanc-
nel power by about 1 dB, but still meets the 6s TTA ing the security of the SBAS augmentation service and
requirement. promoting its applications in the felds of safe-of-life,
such as aviation, navigation, and high-speed train. How-
Te simulation results of SBAS To study the impact of ever, there are still many problems and challenges to be
SBAS authentication on the original SBAS service, the addressed in the authentication of SBAS.
simulation trials were implemented by Fernandez-Her- In terms of system design, the SBAS signal authenti-
nandez et al. (2014, 2018). Te simulations with L1 and cation improves the security of SBAS service, but may