Collision-aware adaptive frequency hopping for anti-jamming backscatter communications
Abstract
This paper develops a lightweight collision-aware adaptive frequency-hopping (FH) framework for anti-jamming backscatter communications. The considered model explicitly captures the cascaded reader–tag–reader propagation path and activates jammer interference only when legitimate and jamming frequencies collide. The resulting hopping problem is formulated as online collision minimization and addressed using an importance-weighted multiplicative update with persistent exploration, requiring only O(M) computation, and O(M) storage per decision without neural-network inference or channel state information (CSI) as a policy input. Monte Carlo evaluation considers uniform, stationary nonuniform, and slowly varying jammer occupancy, with paired realizations and 95% confidence intervals. At M = 16 and Pj = 6 dBm under stationary nonuniform jamming, the proposed method achieves a collision probability of 0.0668 ± 0.0074, an achievable rate of 0.451 ± 0.005 Mbps, a bit error rate (BER) of (2.647 ± 0.021) × 10-1, and an outage probability of 0.9789 ± 0.0010, while requiring 81.67 ± 6.96 µs per decision. Although EXP3, Greedy, and deep Q-network (DQN) attain comparable or lower collision rates, DQN incurs substantially higher runtime. These results demonstrate a favorable tradeoff between collision-aware adaptation, communication performance, and implementation cost for resource-constrained backscatter internet of things (IoT) systems.
Keywords
Anti-jamming; Backscatter communication; Collision-aware frequency hopping; Lightweight adaptive control; Online bandit learning
Full Text:
PDFDOI: https://doi.org/10.11591/eei.v15i6.13022
Refbacks
- There are currently no refbacks.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Bulletin of Electrical Engineering and Informatics (BEEI)
ISSN: 2089-3191
,
e-ISSN: 2302-9285
This journal is published by the
Institute of Advanced Engineering and Science (IAES)
.