Abstract :Space Division Multiplexing-Elastic Optical Networks (SDM-EONs) have emerged as a viable solution to address the exponential growth in data traffic by enhancing spectral efficiency and network scalability. However, spectrum allocation in SDM-EONs presents significant challenges, including spectral fragmentation, latency overhead, and security vulnerabilities. Traditional spectrum allocation methods, such as First Fit (FF) and Machine Learning (ML)-based techniques, fail to effectively integrate security constraints into the allocation process, making networks susceptible to attacks such as eavesdropping, jamming, and route hijacking. This paper introduces a Heuristic Secure Spectrum Allocation (HSSA) algorithm, which employs a multi-metric optimization framework incorporating attack probability, network reliability, and spectrum availability to enhance security-aware spectrum assignment. The proposed method utilizes a modified Dijkstra’s algorithm to compute optimal paths with a security-centric weight function, ensuring minimal fragmentation and efficient spectrum utilization. Extensive simulations on USNET and COST239 network topologies validate the efficiency of HSSA, demonstrating a 95% spectrum utilization rate, 30 ms latency, and enhanced security robustness compared to conventional approaches. The results substantiate the efficacy of HSSA in mitigating spectral inefficiencies and cyber threats while maintaining high resource utilization. Future research will focus on integrating AI-driven dynamic spectrum adaptation, cryptographic security enhancements, and energy-efficient spectrum assignment strategies to further improve SDM-EON performance