Biweekly Summary - April 4, 2025

In this biweekly update, we focus on two aspects. The first one is the coexistence of V2X technologies, and the second one is AI-based resource allocation. 

LTE-V2X and NR-V2X Comparison

From the previous research, we found out the reference of studies comparing the performance of LTE-V2X and NR-V2X is limited. Therefore, we designed and implemented the initial simulation environment to compare LTE-V2X and NR-V2X performance.

We use NS-3 to simulate the  LTE-V2X and 5G-LENA for NR-V2X. We assume to have the simulation result similar to the reference paper we used. We plan to evaluate and compare the Packet Delivery Ratio (PDR), End-to-End Latency, and Throughput for both technologies.

We are currently working on tuning the simulation parameters. However, we have encountered some issues during the process. It might take one or two more days to finalize the parameter tuning. Then we will run the baseline simulations, visualize PDR and got the latency variations for different cases. We will show our result in our presentation and mentioned it in our final report.

This simulation will provide a side-by-side evaluation of LTE-V2X and NR-V2X. We will mention the strengths and limitations for the realistic traffic conditions. The outcome will help guide informed decisions on protocol selection for safety-critical versus data-intensive V2X applications.

Study on Coexistence

• Paper: “Hybrid V2X Communication Approach using WiFi and 4G Connections”

It's become more likely that different technologies will share the same radio specrtrum or frequency in real life. We've read a paper that explores the coexistence challenges and possible performance trade-offs in these scenarios when 802.11p and LTE-V2X operate at the same time.

1.Background on Coexistence

802.11p uses CSMA/CA-based Enhanced Distributed Channel Access (EDCA), and it's a contention-based protocol. While, LTE-V2X uses Semi-Persistent Scheduling (SPS), which means its resources are reserved ahead of time by the vehicle. These fundamental differences between two protocols in channel access methods raise concerns, particularly when both technologies transmit data on the same frequency.

2.Key Challenges of Coexistence

Theoretically, there are several challenges when two protocols coexistence and operate at the same time. The first is lack of coordination, which means the two technologies do not share a spectrum coordination mechanism. Second challenge is interference issues. Adjacent-channel or co-channel interference may be caused by concurrent transmissions. Especially, 802.11p will be affected. Finally, asymmetrical impact could be a challenge too, because LTE-V2X’s SPS protocol gives it more robust channel access. it often overshadows the contention-based 802.11p.

3. Field Test Findings on Coexistence

To see if the prediction is right, the study performed field tests under NLOS conditions (Scenario 2) with both technologies using the same setups: 5 dBm transmission power, 400 bytes packet size and 60–90 meters inter-vehicle gaps.

4. Results of the Test 

The test result showed that when coexisting with LTE-V2X, IEEE 802.11p's Latency increased from 7 ms to 11 ms and packet Delivery Ratio (PDR) dropped from 67% to 52%. While LTE-V2X has a minimal latency increase, and its PDR slightly decreased from nearly 100% to 89%.

These results proved the unfair channel access where LTE-V2X dominates, resulting in notable degradation in 802.11p’s performance.

5. Related Work & Mitigation Strategies

The paper also suggests a few coexistence strategies to mitigate the degradation. 

• Channel Separation: Allocating different channels for each technology. But the cost of this method is reducing the total usable spectrum.

• Cross-technology Relaying: Using dual-interface vehicles to bridge protocols.

• Preamble Insertion: Embedding 802.11p preambles into LTE-V2X frames to enhance mutual awareness.

However, most of these have only been tested in simulations and not validated through field tests by the writer.

AI-driven Resource Allocation

AI-driven resource allocation can improve the performance of V2X technologies by optimizing the allocation of scarce radio resources. By doing this, it can reduce interference and improve communication reliability and latency. There are several ways for AI to contribute to V2X performance:

1. AI algorithms can allocate resources based on real time network conditions. For example, deep reinforcement learning techniques can enable V2X systems to autonomously choose optimal radio resources based on channel state information (CSI). It can minimize interference between vehicles. Also, it can maintain the required reliability and latency for V2X communications.

2. AI models can learn patterns in vehicular environments, such as vehicle density and mobility patterns, to predict network congestion and proactively manage bandwidth allocation. So we can have better decisions based on the distribution of resources. This can ensure lower latency and higher packet delivery success rates.

3. AI can accurately predict network conditions and adjust resource allocations. So it can improve reliability. Reliability is important for vehicular platooning which requires low latency and high reliability.

4. AI techniques can scale with network size, adapting to the increasing number of connected vehicles. As the density of vehicles changes, AI models can adjust resource allocations in real time to maintain optimal network performance.