Abstract
For D2D-based vehicular communication networks, how to effectively and quickly allocate frequency resources is a big challenge. In this paper, a resource allocation scheme based on immune algorithm for D2D-based vehicular communication networks is proposed. Firstly, the D2D-based vehicular communication system model is described. Then the resource allocation issue in D2D-based vehicular communication networks is formulated and analyzed. Next, an efficient resource allocation scheme based on immune algorithm is presented. Finally, the analysis and simulation results validate the high efficiency of the proposed scheme compared with other algorithms. This result can provide the guidance for the deployment of VUE communication system.
Introduction
The rapid proliferation of vehicles on modern roadways has catalyzed a surge in demand for high-speed data services and low-latency communication, essential for both safety-critical applications and passenger infotainment. Device-to-Device (D2D) communication has emerged as a pivotal technology in this landscape, enabling Vehicle-to-Vehicle (V2V) links that bypass traditional base station routing. This direct communication paradigm reduces latency and offloads traffic from the core network.
However, integrating D2D links into existing cellular frameworks introduces significant technical hurdles, primarily centered on interference management. In these heterogeneous networks, Vehicle User Equipments (VUEs) often reuse the same frequency spectrum as Cellular User Equipments (CUEs). The high mobility of vehicles causes the network topology to change dynamically, making the optimal assignment of frequency resources an NP-hard problem. Traditional optimization techniques often fail to meet the real-time constraints of vehicular environments due to high computational complexity.
System Model and Problem Formulation
The proposed scheme considers a single-cell vehicular network where multiple VUE pairs and CUEs coexist. The goal is to maximize the total system throughput while satisfying the minimum Signal-to-Interference-plus-Noise Ratio (SINR) requirements for both cellular and D2D links.
As illustrated in the system model, the network consists of a central Base Station (BS), CUEs, and VUE pairs (transmitters and receivers). The interference environment is complex: CUEs interfere with VUE receivers, and VUE transmitters interfere with both the BS and other VUE receivers. Effective resource allocation must navigate these cross-tier and co-tier interferences to maintain link reliability.
The Immune Algorithm Approach
To address the resource allocation challenge, this research proposes a bio-inspired approach based on the Immune Algorithm (IA). The IA mimics the adaptive and self-organizing nature of biological immune systems to find near-optimal solutions efficiently.
Algorithmic Mapping
In this framework, the optimization problem is mapped to immune system components:
- Antigens: Represent the optimization objective (maximizing system throughput).
- Antibodies: Represent potential resource allocation solutions, encoded as M-dimensional vectors.
- Affinity: A fitness score for each antibody, calculated based on the resulting total system throughput.
Dynamic Adaptive Cloning
A key innovation in this scheme is the dynamic adaptive clone operator. Unlike traditional evolutionary algorithms that use static cloning rates, this operator adjusts the number of clones for an antibody based on its affinity. High-performing solutions (high affinity) are cloned more frequently, accelerating the search for the global optimum.
Procedural Steps
- Encoding: Solutions are initialized as random vectors representing resource assignments.
- Evaluation: The SINR and total throughput for each antibody are calculated to determine affinity.
- Immune Operations: The population undergoes selection, adaptive cloning, and mutation. Mutation introduces diversity, preventing the algorithm from converging prematurely on local optima.
- Inhibition and Update: Low-quality solutions are pruned (inhibited), and new random antibodies are introduced to maintain population vitality.
Performance Evaluation
The effectiveness of the IA-based scheme was evaluated using Monte Carlo simulations. The simulation environment featured a 600m radius cell with a straight road passing 200m from the BS. Traffic was modeled using a spatial Poisson process, involving 4 CUEs and 8 VUE pairs with speeds ranging from 40 km/h to 90 km/h.
Key Findings
- Throughput Performance: The IA achieved system throughput remarkably close to the Exhaustive Algorithm (EA), which represents the theoretical upper bound. It outperformed the Harmony Search Algorithm (HSA) by approximately 2 Mbps and the Random Algorithm (RA) by 4 Mbps.
- Computational Efficiency: One of the most significant advantages of the IA is its speed. The computational complexity was found to be roughly half that of the HSA, making it highly suitable for real-time vehicular applications where decision-making must occur in milliseconds.
- Mobility Robustness: While increased vehicle speed generally degrades performance due to rapid channel changes, the IA maintained a consistent performance lead. For instance, at higher speeds, the IA remained approximately 3.4% more efficient than the HSA.
- Convergence Speed: The IA demonstrated rapid convergence, reaching stable high-performance states in fewer iterations than competing intelligent algorithms.
Conclusion
The proposed Immune Algorithm-based resource allocation scheme offers a robust and efficient solution for D2D-based vehicular networks. By leveraging adaptive cloning and immune memory, the algorithm achieves near-optimal throughput with significantly reduced computational overhead compared to traditional methods. While the current study utilizes a simplified single-cell model, it provides a strong foundation for future research into multi-cell urban environments and more complex fading models. This work offers valuable guidance for the practical deployment of high-performance VUE communication systems.
Summary updated