ATRAS: An Adaptive Traffic- and Security-Aware Resource Allocation Scheme for eMBB, URLLC, and mMTC Coexistence in 5G Heterogeneous Networks

Abstract

Most existing 5G resource allocation schemes address only one or two of the eMBB, URLLC, and mMTC use cases, apply static, service-level prioritization that favours one class over the others, and assume every UE reports its traffic class and congestion state truthfully, an assumption that leaves adaptive, priority-based allocation open to UEs that spoof a higher-priority QoS class or flood the network with false delay signals to trigger unwarranted resource pre-emption. This paper proposes the Adaptive Traffic- and Security-Aware Resource Allocation Scheme (ATRAS), an inter-use-case scheme that assigns resource blocks to eMBB, URLLC, and mMTC by priority weight, dynamically pre-empts a bounded share from the lowest-priority use case only when the URLLC delay budget is genuinely at risk, and screens every pre-emption trigger against the requesting UE's authenticated QoS class and its recent triggering history before acting, rejecting requests that look spoofed or flood-induced. ATRAS was evaluated through MATLAB-based simulation under a mission-critical industrial traffic mix and benchmarked against two recent eMBB–URLLC schemes, neither security-aware, using average delay, packet drop ratio, nor a new security detection rate across loads of 25 to 200 UEs. Results show ATRAS sustains lower delay and drop ratio than both benchmarks, with the margin reaching roughly 32% and 42% at 200 UEs, while correctly detecting over 90% of spoofed or flood-induced pre-emption attempts at every load tested.