The eagerly anticipated 5G use case of autonomous vehicles is set to revolutionize transportation, promising new services such as pollution monitoring, predictive diagnostics, congestion detection, and enhanced city mobility, alongside more reliable infotainment. Cloud architecture is the key to realizing these innovations, but significant technical challenges must be addressed to make cars truly smart. So, now let us see How Can OpenStack and ETSI Open-Source MANO Revolutionize Edge Architecture for Self-Driving Cars along with Accurate Best wireless site survey software, site survey tools for wireless networks & Indoor cellular coverage walk testing tool and Accurate LTE RF drive test tools in telecom & RF drive test software in telecom in detail.
Key Challenges:
Low Latency Requirements: Vehicles need simultaneous data transmission between onboard units (OBUs) and cloud or data centers.
Rapid Data Processing: Raw data generated by vehicles must be quickly processed to provide actionable insights.
Intermediate Data Layer: An intermediate layer is needed to cache raw data from multiple vehicle sensors for immediate insight generation.
Researchers at Cartagena and Murcia universities in Spain proposed a solution called “surrogates” to address these challenges by virtualizing vehicle OBUs using OpenStack and ETSI Open Source Mano (OSM). Their proof-of-concept involves a multi-access edge computing (MEC) layer that acts as a middle layer between the cloud and vehicle OBUs.
Solution Overview:
The solution virtualizes the functionality of physical OBUs into virtual network functions (VNFs) hosted on the network edge (e.g., roadside units, cellular base stations, WiFi access points). These VNFs perform data pre-processing tasks such as analysis, pattern recognition, and data aggregation, managed by NFV at the edge and then pushed to the cloud for high-volume global big data analysis. Processed insights are accessed by OBUs from vOBUs or a data-analytics module.
By moving computing resources from physical OBUs to the edge layer, the VNFs can be accessed simultaneously by multiple OBUs. This setup reduces data latency and transmission time by 50% and frees up wireless network bandwidth for numerous vehicles and devices in the network.
Technical Implementation:
The proposed solution requires two NFV clouds: a central cloud and an edge infrastructure. Built using OpenStack and ETSI Open Source Mano (OSM), the setup gathers and processes data from vehicle OBUs. OpenStack (version Pike) acts as a virtual infrastructure manager (VIM) for both infrastructures, while OSM (version 4) orchestrates VNFs at the central cloud. The edge infrastructure domain, equipped with MEC capabilities, manages vOBU VNFs using OpenStack, which communicate with MANO, the OBU manager, and data analytics modules.
The OBU manager, a VNF deployed at the edge, acts as an intermediary for physical OBUs and their virtual counterparts, communicating with OSM for allocation, scaling, and status reports. When new requests arise or new vehicles join the network, scaling and allocation of vOBUs occur.
Monitoring is handled by a module deployed at the physical OBU and OpenStack environment at the edge, using Grafana for monitoring solutions. Data analytics modules respond to analysis requests from edge VNFs, providing new or cached insights.
Proof-of-Concept Evaluation:
The PoC was tested under the 5GINFIRE project testbed setup with three key experiments:
Evaluating platform operation with different cache success rates in the vOBU.
Studying the ratio of request losses.
Analyzing system scalability and the duration of new-vehicle registration and vOBU instance allocation.
Validation through a monitoring service using Grafana and an Android app demonstrated the proposal’s efficacy, as researchers concluded.
Conclusion:
This proposed solution illustrates how offloading device functions to the edge domain can benefit not only vehicular services but also other use cases like industrial IoT and smart city operations. Leveraging leading open infrastructure projects such as OpenStack and OSM as MANO shows that such solutions are feasible with open-source projects, offering the added benefit of controlled infrastructure development costs. OpenStack and OSM bring robust solutions to support 5G use cases on a larger scale, paving the way for a smarter, more connected future. Also read similar articles from here.