Distributed IP Allocation via DHT for Local Data Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current Evolved Packet Core (EPC) network architecture is not suited for efficiently processing locally terminated data, which is increasingly demanded by enterprise and industry segments, as it relies on a hierarchy of Serving Gateways and Packet Data Network Gateways that are not optimized for local data processing and IP address allocation.
Innovation Solution
A distributed IP address allocation mechanism is implemented using a distributed hash table (DHT) that allows packet processing units (PPUs) in a communications network to independently allocate IP addresses from a shared pool, with each PPU maintaining a distributed block allocation table and address allocation tables, enabling them to act as independent IP address allocation servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a centralized PGW architecture is used for IP address allocation, then network management is simplified, but the system cannot efficiently process locally terminated data and introduces single point of failure risks
Solution Approach 1:
The patent segments the centralized PGW functionality into multiple distributed PPUs, each capable of independent IP address allocation. The DHT further segments the address allocation tables across multiple nodes, allowing locally terminated data to be processed by any PPU in the distributed network without requiring centralized coordination.
Solution Approach 2:
Each PPU in the distributed architecture maintains local copies of address allocation information and can independently allocate IP addresses without querying a centralized server. The system provides self-service through local caching and autonomous decision-making at each PPU node.
2Reliability
If IP address allocation is centralized in a single server, then address management is simplified, but the system lacks redundancy and self-healing capabilities
Solution Approach 1:
The address allocation table is segmented and distributed across multiple PPUs using the DHT structure. Each PPU manages a portion of the address space, and multiple replicas of allocation information are maintained across the network, providing redundancy and eliminating single points of failure.
Solution Approach 2:
The system implements feedback mechanisms where PPUs periodically exchange allocation status information through the DHT, allowing the network to detect and recover from failures automatically. When a PPU fails, other PPUs detect the failure through missing feedback and can take over its address allocation responsibilities.
3Object-affected harmful factors
If all IP address allocation requests are handled by centralized PGWs, then address allocation is straightforward, but network delay increases and security is compromised
Solution Approach 1:
The patent implements local quality by allowing each PPU to make autonomous IP address allocation decisions based on local cached information from the DHT. This local decision-making eliminates the need for all requests to traverse to centralized PGWs, reducing network delay and improving security by enabling local data termination.
Solution Approach 2:
PPUs perform preliminary actions by maintaining local caches of address allocation tables in advance. This allows them to quickly allocate addresses without real-time queries to centralized servers, reducing delay while the periodic updates to the DHT ensure data consistency.
4Productivity
If multiple PPUs independently allocate IP addresses from a shared pool, then locally terminated data processing is enabled, but coordinate management becomes complex
Solution Approach 1:
The DHT provides a universal coordination mechanism that all PPUs use to manage the shared IP address pool. This single distributed data structure serves multiple functions: tracking allocated addresses, determining PPU responsibilities, and enabling autonomous allocation decisions across all nodes simultaneously.
Solution Approach 2:
The DHT acts as an intermediary between multiple independent PPUs, mediating their access to the shared address pool. Each PPU queries the DHT to determine which address ranges it can allocate, and the DHT coordinates these allocations without requiring direct PPU-to-PPU communication or complex distributed locking protocols.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
A non-transitory computer readable medium has instructions stored therein to be executed by a packet processing unit (PPU) in a communications network to allow a plurality of PPUs in the communications network to independently allocate IP addresses from a shared pool of IP addresses. The plurality of PPUs in the communications network collectively maintain a distributed hash table (DHT). The DHT stores a distributed block allocation table (DBAT) and a plurality of distributed address allocation tables (DAATs). The DBAT includes an entry for each of a plurality of blocks of IP addresses to indicate which of the plurality of blocks are allocated to which of the plurality of PPUs. Each of the plurality of DAATs corresponds to one of the blocks in the DBAT and includes an entry for each of the IP addresses in that block to indicate which of the IP addresses in that block are allocated to which user entities (UEs).