Received: 2 March 2017 Revised: 3 December 2017 Accepted: 12 December 2017
Adaptive frequency-hopping schemes for CR-based multi-link
Irmak Aykin Marwan Krunz Yong Xiao
Department of Electrical and Computer
Engineering, University of Arizona, Tucson, AZ
Irmak Aykin, Department of Electrical and
Computer Engineering, University of Arizona,
Tucson, AZ 85721, USA.
NSF, Grant/Award Number: CNS-1563655
and CNS-1513649; Broadband Wireless
Access & Applications Center (BWAC)
In this paper, we study two dynamic frequency hopping (DFH)–based interference mitigation
approaches for satellite communications. These techniques exploit the sensing capabilities of
a cognitive radio to predict future interference on the upcoming frequency hops. We con-
sider a topology where multiple low Earth orbit satellites transmit packets to a common geo-
stationary equatorial orbit satellite. The FH sequence of each low Earth orbit–geostationary
equatorial orbit link is adjusted according to the outcome of out-of-band proactive sensing
scheme, performed by a cognitive radio module in the geostationary equatorial orbit satellite.
On the basis of sensing results, new frequency assignments are made for the upcoming slots,
taking into account the transmit powers, achievable rates, and overhead of modifying the FH
sequences. In addition, we ensure that all satellite links are assigned channels such that their
minimum signal-to-interference-plus-noise ratio requirements are met, if such an assignment is
possible. We formulate two multi-objective optimization problems: DFH-Power and DFH-Rate.
Discrete-time Markov chain analysis is used to predict future channel conditions, where the
number of states are inferred using k-means clustering, and the state transition probabilities
are computed using maximum likelihood estimation. Finally, simulation results are presented to
evaluate the effects of different system parameters on the performance of the proposed designs.
cognitive radio, dynamic frequency hopping, Markov models, maximum likelihood estimation,
out-of-band sensing, satellite communications
RF interferencecan significantly degrade the performance of satellite networks. To cope with that, spreading systems are often used as the first “line
of defense" against active (jamming) and passive (eavesdropping) attacks.
In direct sequence (DS) spread spectrum, the signal is passed through a
spreading function and distributed over the entire band at once. On the other hand, in frequency-hopping (FH) spread spectrum, the carrier fre-
quency of the signal is switched between a set of frequencies using a pseudorandom noise sequence. Even though both approaches tend to work
well against eavesdroppers, FH spread spectrum is more robust to interferers than DS spread spectrum. Some examples of satellite systems that
use FH include the Advanced Extremely High Frequency
and the Military Strategic and Tactical Relay (MILSTAR) satellite systems.
Indeed, FH is
extensively used in military satellite systems to secure communications, especially under adversarial conditions.
Although traditional FH where
the carrier frequencies are changed using a fixed pseudorandom noise sequence tends to work well under random interference, it performs poorly
in the presence of persistent interference over specific channels.
For example, if a smart eavesdropper learns parts of the fixed FH sequence, it can
persistently target certain frequencies in certain time slots.
Onewaytoaddressthe limitations of the static FH approach is to use dynamic frequency hopping (DFH),inwhichtheFH sequence can be adjusted
accordingto channel quality and interference conditions. DFH was studied in Hu et al
toensure a certain quality of service (QoS) in wirelessregional
An abridged version of this paper was presented at the IEEE Globecom 2016 Conference, Washington, DC, December 4-8, 2016.
Int J Satell Commun Network. 2018;36:315–331. wileyonlinelibrary.com/journal/sat Copyright © 2018 John Wiley & Sons, Ltd. 315