Virtual Network Function State Management in Mobile Edge Computing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In mobile edge computing, existing solutions face challenges in maintaining low latency and efficient computation offloading for user equipment (UE) with uncertain or unpredictable trajectories, especially in scenarios where network function (NF) migration is complex due to autonomous or semi-autonomous motion.
Innovation Solution
A method and system that allow UE to offload computational tasks by transmitting network function images and state information, enabling seamless transition between MEC resources without direct state transfer between control entities, allowing newly created or allocated functions to treat state information as stateless, and maintaining it through UE transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If the network function is migrated to follow the UE trajectory, then the communication latency is reduced, but the complexity of state transfer between network functions increases
Solution Approach 1:
The patent introduces the UE as an intermediary carrier for state information. Instead of directly transferring state between network functions, the state is encapsulated in NF images that are transmitted through the UE. This mediator approach simplifies the transfer mechanism while maintaining low latency by keeping the state close to the moving UE.
Solution Approach 2:
The patent creates copies of the network function state in the form of NF images. These images are transmitted to and stored at the UE, allowing the network function to be reconstructed at new network nodes without complex direct state transfer. The copying mechanism enables efficient state migration while reducing transfer complexity.
2Loss of time
If multiple network functions are deployed in different network locations to support UE mobility, then the latency is reduced, but the device complexity and management overhead increase
Solution Approach 1:
The patent makes the NF image format universal and platform-independent, allowing the same NF image to be executed on different network nodes throughout the network. This multi-functionality enables a single NF image to serve multiple locations, reducing the need for location-specific function instances and simplifying management overhead while maintaining low latency.
Solution Approach 2:
The patent implements dynamic NF instantiation and migration based on UE movement. Network functions are created, moved, and deactivated dynamically as the UE traverses the network, rather than maintaining static function deployments. This dynamic approach optimizes latency by keeping functions close to the UE while reducing management complexity through automated lifecycle management.
3Loss of time
If the network predicts and pre-positions network functions along the UE trajectory, then the latency is further reduced, but the reliability decreases when the UE trajectory is uncertain or autonomous
Solution Approach 1:
The patent performs preliminary actions by pre-processing and encapsulating the network function state into NF images before migration. The NF images are prepared in advance with all necessary state information, allowing quick deployment at new locations without complex real-time state transfer. This preliminary preparation maintains reliability even when trajectory prediction is uncertain.
Solution Approach 2:
The patent enables the NF images to be self-contained and self-sufficient, carrying all necessary state information within them. When migrated to new network nodes, the NF images can be independently instantiated without requiring complex coordination or validation from the network, making the system more reliable in the face of uncertain UE trajectories.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
In an embodiment a method is provided for processing a computational task using a network node available on a network processing. The method including the network node performing the steps of processing a computational task, and transmitting at least one snapshot of state data related to the computational task to a UE associated with that computational task. The at least one snapshot of state data enables the UE to resume processing of the computational task from a snapshot point represented by the state data.